Predicting Acute Respiratory Infection Risk in Under–Five Children Using Machine Learning: Evidence from Bangladesh

Predicting Acute Respiratory Infection Risk in Under–Five Children Using Machine Learning: Evidence from Bangladesh | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Predicting Acute Respiratory Infection Risk in Under–Five Children Using Machine Learning: Evidence from Bangladesh Samrat Kumar Dev Sharma, Md. Yusuf Hossain Ador, Md. Rokunuzzaman, and 4 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7081278/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Background Children’s immune systems are particularly vulnerable to infections. In developing countries, malnutrition, diarrheal diseases, and acute respiratory infections (ARI) remain leading causes of illness and mortality among children. This study applied multiple machine learning (ML) techniques to identify key risk factors associated with ARI symptoms in Bangladesh. Methods Secondary data from the Bangladesh Demographic and Health Survey (BDHS) 2022 were analyzed to assess ARI risk factors. Feature selection was conducted using the SHAP algorithm, and six ML classifiers were trained and evaluated with 10-fold cross-validation. Model performance was measured using accuracy, sensitivity, specificity, precision, F1-score, G-mean, and ROC AUC. Results Among the classifiers, the Decision Tree (DT) model achieved the highest performance across several metrics, with accuracy, sensitivity, specificity, precision, F1-score, and G-mean all at 0.81, and an ROC AUC of 0.88. Random Forest (RF) and AdaBoost (AdaB) also demonstrated strong performance, with RF showing an accuracy of 0.79 and ROC AUC of 0.89, and AdaB achieving an accuracy of 0.78 and ROC AUC of 0.86. Key predictive features included fever, age groups, geographic division, and area of residence. Conclusions The Decision Tree classifier outperformed other ML models in predicting ARI risk among children under five in Bangladesh, closely followed by Random Forest and AdaBoost. These findings highlight the potential of ML approaches to support targeted interventions by identifying critical risk factors. Government strategies should focus on early detection, improved treatment of fever, and enhanced household conditions to mitigate ARI risk in young children. ARI children risk factors machine learning fever geographic division Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 1. Introduction Children's immune systems are particularly susceptible to illness. Malnutrition, diarrheal illnesses, and acute respiratory infections (ARI) are the leading causes of child sickness and death in developing countries (1). As to the Global Health Observatory's (GHO) 2016 report, the leading causes of death for children under the age of five were preterm birth complications and acute respiratory infections (ARI) (2). ARI can be caused by a variety of pathogens, including viruses, bacteria, and, sometimes, fungi (3). Pneumonia is one of the most prevalent and severe forms of ARI among children under five (4). In 2019, the mortality rate for children younger than five in Sub-Saharan Africa was 78 per 1000 live births, double that of the world and at least 16 times higher than the average of high-income nations (1). The death rate among children under the age of five in Africa was nearly eight times that of Europe. Furthermore, air pollution disproportionately affects children living in low- and middle-income nations (5). In countries with high incomes, breakthroughs in medical care, access to safe sanitation and water, and enhancements in all aspects of life have lowered the infant mortality rate to 5 per one thousand live births (6). Sub-Saharan Africa and Southeast Asia comprised the vast majority of ARI mortality in children under the age of five. ARIs continue to be the most significant burden in developing countries, including Ethiopia and Bangladesh. According to the Bangladesh Demographic and Health Survey (BDHS), ARI causes around 25% of all fatalities among children under the age of five in Bangladesh each year. In accordance to the Bangladesh DHS, 2017-18 - Final Report, 80% of households in Bangladesh cook with solid fuels (coal/lignite, charcoal, wood, straw/shrubs/grass, crops, and animal dung), while 20% use clean fuels (electric power and petroleum-based gas/natural gas/biogas). Furthermore, the primary cause of mortality and morbidity in Bangladesh for diarrhea has been effectively controlled, while the associated risk factors for ARI have been growing by the day (7). Bangladesh is an LMIC, with about 166 million people (63%) residing in the countryside (8). Nevertheless, the precise size of ARI, which is currently increasing on a massive scale in Bangladesh, remains unknown. Unlike diarrhea or acute malnutrition, ARI has no acceptable standards, making it impossible to assess case treatment excellence using established criteria. Several research investigations found substantial associations between developing ARI and environmental risk characteristics, such as smoke created by indoor cooking, various types of air pollution from the outdoors, passive smoking, and crowding (9,10). These hazards in children under the age of five produce a variety of severe difficulties, including low birth weight, starvation, measles, pneumonia, and problems with breastfeeding (9). Furthermore, various types of fuel for cooking, poor sanitation conditions, mother education rate, sufficient medication for bowel parasitic organism, place of residence, mother's and child's body mass index (BMI), and financial status index are all possible risk factors for pneumonia/ARI in nations that are developing such as Bangladesh (11,12). Multiple research investigations on healthcare-seeking behavior for signs of ARIs among children under the age of five were undertaken in Bangladesh, Nigeria, East Africa, and rural Kenya (2,13), Ghana, and sub-Saharan Africa (4). Earlier studies used linear and non-linear regression models to investigate the determinants of ARIs in children under the age of five. As long as the researcher is aware, there have been a few previous studies that used algorithmic machine learning to predict ARIs in children younger than five using environmental indicators (5). Previously, conventional regression models were employed to undertake substantial research on the socioeconomic and demographic characteristics related to acute respiratory infections in Ethiopia (13). Recent developments in machine learning (ML) provide effective methods for analyzing intricate, high-dimensional health data and enhancing disease risk assessment. However, despite their potential, ML models have not been widely applied to evaluate ARI risk in children under five in Bangladesh. The BDHS 2022 dataset presents a chance to implement ML techniques to investigate the interactions between clinical, demographic, and environmental factors. This research utilizes various ML classifiers and employs SHapley Additive exPlanations (SHAP) for analyzing feature importance to forecast ARI risk in young children. By evaluating the effectiveness of different models, the study seeks to enhance predictive accuracy and reveal hidden patterns within the data. The results will offer important insights for policymakers and healthcare practitioners to create targeted strategies for the early detection and prevention of ARI, ultimately contributing to a reduction in morbidity and mortality rates among children in Bangladesh. 2. Methodology This study is based on a cross-sectional secondary data analysis using the Bangladesh Demographic and Health Survey (BDHS) 2022, which provides nationally representative data on demographic, reproductive, and health indicators. The BDHS employed a two-stage stratified sampling strategy. Initially, 675 enumeration areas (EAs) were chosen, featuring 438 from rural zones and 237 from urban environments, based on probability proportional to size. In the second stage, 30 households were selected systematically from each EA, resulting in a total sample of 30,078 households (14). The study population was restricted to children under five years of age. Children were included in the analysis if they had symptoms of acute respiratory infection (ARI) within the two weeks preceding the survey, as identified by caregiver-reported short, rapid breathing and or difficult breathing that is chest-related. For the analysis of care-seeking behavior, only children with reported ARI symptoms were included. Care-seeking was defined as seeking advice or treatment from qualified healthcare sources, including public sector providers, private medical facilities, or nongovernmental organizations (NGOs). Children for whom care was sought only from traditional practitioners were excluded from this portion of the analysis, in accordance with BDHS 2022 definitions. After applying these criteria, the final analytic sample consisted of 21,899 children. The overall workflow shows in Fig. 1 . 2.1 Study variables Acute respiratory infection (ARI) was identifie based on caregiver reports of coughing symptoms occuring within the two weeks prior to the survey. Follwing the BDHS guidelines, the outcome variable was binary, categoried as (0 = refer no ARI and 1 = refer yes ARI). A set of independent variables were considered in the analysis based on previous literature review (1,2,4,13). These included age in months (< 6, 6–11, 12–23, 24–35, 36–47, and 48–59), sex of child (male, female), breastfeeding status (yes or no), recent history of diarrhea and fever (yes or no), and body mass index (BMI), categorized using standard thresholds (underweight, normal, overweight, obese). Socio-demographic variables included place of residence (urban or rural), administrative division (Barisal, Chattogram, Dhaka, Khulna, Mymensingh, Rajshahi, Rangpur, and Sylhet), and household wealth index (poorest, middle, richest). Maternal characteristics included mother’s education level (none/pre-primary, primary, secondary, or higher), presence of health difficulties (yes or no), and whether the child had any reported disability (yes or no). 2.2 Data preprocessing This study, the data preparation process was to ensure the quality and integrity of the dataset prior to analysis. Different key variables including ARI status, age, sex, diarrhea, fever, BMI, area, division, wealth index, mother’s education, mother’s difficulties had no missing values. However, a substantial proportion of missing data was seen in breastfeeding status (42.22%) and child disability status (37.13%). These missing values were addressed using mode imputation. Categorical variables such as age, sex, breastfeeding status, diarrhea, fever, area, division, child disability, mother’s education, mother’s health difficulties, wealth index, and BMI category were converted into dummy variables using one-hot encoding. To avoid multicollinearity, the first category of each variable was dropped during encoding (15). The new dummy variables were combined with the dataset, and the original categorical columns were removed. After encoding, the dataset was randomly split into 80% for training and 20% for testing. To resolve the issue of class imbalance in the ARI outcome, the Synthetic Minority Over-sampling Technique (SMOTE) was applied. SMOTE generated synthetic instances of the minority class (children with ARI symptoms), resulting in a balanced dataset that improved the stability and generalizability of predictive models (16). Following preprocessing and resampling, feature selection was conducted using a Random Forest model combined with SHAP (SHapley Additive exPlanations) values (17). SHAP values were used to measure how much each feature contributed to the model's prediction. Features with higher SHAP values were considered important and kept for further analysis, while less important ones were removed. This method helped select the most relevant variables and made the model easier to understand and more correct (Fig. 2 ). Machine learning algorithms were employed to predict acute respiratory infections (ARI) and identify their contributing factors. Data processing and analysis were carried out in Python, utilizing libraries such as Pandas, Scikit-learn, Imbalanced-learn, NumPy, and Seaborn for tasks including data cleaning, transformation, model training, and evaluation. A predictive model was developed to estimate the risk of ARI and determine the key associated factors. 2.3 Machine Learning Models To predict acute respiratory infections (ARI) and identify associated risk factors, several machine learning algorithms were applied. These models were selected based on their effectiveness in classification tasks, diversity in learning mechanisms, and interpretability. Random Forest is a strong ensemble technique that builds several decision trees by employing random samples of the data and features. Final predictions are made through majority voting (for classification) or averaging (for regression). This approach reduces overfitting and improves generalization (Liaw, Wiener and others, 2002). Extremely Randomized Trees (Extra Trees) operate similarly to Random Forests but introduce greater randomness by selecting split thresholds randomly rather than optimizing them. This can result in faster training and improved generalization performance in certain cases (Dietterich, 2000). Adaptive Boosting (AdaBoost) is an ensemble technique that combines multiple weak learners, typically shallow decision trees, into a strong classifier. It works iteratively by assigning higher weights to misclassified instances, enabling the model to focus on more difficult cases in subsequent rounds (Natras, Soja and Schmidt, 2022). Logistic Regression (LR) is a statistical model frequently employed for classifying binary outcomes. It estimates the probability of a class label by applying the logistic function to a linear combination of input features. Logistic regression is appreciated for its simplicity, efficiency, and interpretability (Hosmer Jr, Lemeshow and Sturdivant, 2013). Support Vector Classifier (SVC), a form of support vector machine, finds the optimal hyperplane that separates classes in a high-dimensional space. It is particularly effective in handling non-linear relationships through the use of kernel functions (22). Decision Tree (DT) is a non-parametric model that splits the data into branches based on feature values, forming a tree-like structure. While easy to interpret and visualize, decision trees are susceptible to overfitting, which can be mitigated through pruning or regularization techniques (Dietterich, 2000b). 2.4 Evaluation matrices Model performance was evaluated using accuracy, sensitivity, specificity, precision, F1- score, G-Mean score and ROC(AUC) score. In this framework, TP says true positives, TN is negative, FP stands for false positives, and FN denotes false negatives. 3. Results 3.1 Results of the background characteristics Among the 21,899 children under five, ARI prevalence was highest in the 6–11-month age group (4.3%) and lowest in the 48–59-month group (1.4%). Male children had a higher ARI rate (2.6%) than females (1.8%). Children who had fever (6.3%) or diarrhoea (3.6%) showed significantly higher ARI rates than those without these symptoms. ARI was slightly more common among urban (2.3%) than rural (2.2%) children. Children with disabilities (4.4%) and those whose mothers reported health difficulties (5.5%) had higher ARI prevalence. Maternal education and household wealth showed minimal variation in ARI rates, which remained around 2.2–2.4%. Nutritional status was dominated by underweight children (92.2%), with no ARI reported among overweight children. Regionally, the highest ARI rate was found in Mymensingh (6.3%), while Sylhet had the lowest (0.8%) (Table 1 ). Table 1 Background Characteristics of the Study Populations Variables Frequency N (%) Acuite Respiratory System (ARI) No (%) Yes(%) Age (in Month) 0–5 2291 (10.46) 2212 (96.6) 79 (3.4) 6–11 1452 (6.63) 1390 (95.7) 62 (4.3) 12–23 4408 (20.13) 4295 (97.4) 113 (2.6) 24–35 4513 (20.61) 4411 (97.7) 102 (2.3) 36–47 4696 (21.44) 4626 (98.5) 70 (1.5) 48–59 4539 (20.73) 4476 (98.6) 63 (1.4) Sex (Child) Female 10593 (48.37) 10403 (98.2) 190 (1.8) Male 11306 (51.63) 11007 (97.4) 299 (2.6) Breastfeeding No 566 (2.58) 553 (97.7) 13 (2.3) Yes 21333 (97.42) 20857 (97.8) 476 (2.2) Diarrhoea No 20429 (93.29) 19993 (97.9) 436 (2.1) Yes 1470 (6.71) 1417 (96.4) 53 (3.6) Fever No 16898 (77.16) 16722 (99.0) 176 (1.0) Yes 5001 (22.84) 4688 (93.7) 313 (6.3) Area Rural 17790 (81.24) 17396 (97.8) 394 (2.2) Urban 4109 (18.76) 4014 (97.7) 95 (2.3) Child Disability No 21539 (98.36) 21066 (97.8) 473 (2.2) Yes 360 (1.64) 344 (95.6) 16 (4.4) Mother Education Pre-primary 2354 (10.75) 2308 (98.0) 46 (2.0) Primary 5273 (24.08) 5148 (97.6) 125 (2.4) Secondary 10830 (49.45) 10590 (97.8) 240 (2.2) Higher secondary 3442 (15.72) 3364 (97.7) 78 (2.3) Mother Difficulty No 21589 (98.58) 21117 (97.8) 472 (2.2) Yes 310 (1.42) 293 (94.5) 17 (5.5) Wealth Index Poorest 5408 (24.70) 5288 (97.8) 120 (2.2) Middle 12823 (58.56) 12537 (97.8) 286 (2.2) Richest 3668 (16.75) 3585 (97.7) 83 (2.3) BMI Underweight 20197 (92.23) 19746 (97.8) 451 (2.2) Normal 853 (3.90) 834 (97.8) 19 (2.2) Overweight 817 (3.73) 32 (100.0) 0 (0.0) Obesity 32 (0.15) 798 (97.7) 19 (2.3) Division Barishal 1953 (8.92) 1897 (97.1) 56 (2.9) Chattogram 4547 (20.76) 4457 (98.0) 90 (2.0) Dhaka 4299 (19.63) 4228 (98.3) 71 (1.7) Khulna 3023 (13.80) 2956 (97.8) 67 (2.2) Mymenshing 1293 (5.90) 1212 (93.7) 81 (6.3) Rajshahi 2274 (10.38) 2217 (97.5) 57 (2.5) Rangpur 2609 (11.91) 2557 (98.0) 52 (2.0) Sylhet 1901 (8.68) 1886 (99.2) 15 (0.8) 3.2 Experiments Results The predictive performance of six machine learning classifiers Random Forest (RF), Extra Trees (ET), AdaBoost (AdaB), Logistic Regression (LG), Support Vector Classifier (SVC), and Decision Tree (DT) was evaluated using several metrics including accuracy, sensitivity, specificity, precision, F1-score, ROC AUC, and G-Mean (Table 2 ). Table 2 Performance Metrics score for Classical ML Classifiers Models Accuracy Sensitivity Specificity Precision F1-score ROC AUC G-Mean RF 0.79 0.79 0.79 0.79 0.79 0.89 0.79 ET 0.76 0.73 0.78 0.77 0.75 0.83 0.75 AdaB 0.78 0.78 0.79 0.79 0.78 0.86 0.78 LG 0.75 0.76 0.74 0.75 0.75 0.83 0.75 SVC 0.74 0.74 0.75 0.74 0.74 0.82 0.74 DT 0.81 0.81 0.81 0.81 0.81 0.88 0.81 Among the models, the Decision Tree (DT) achieved the highest accuracy of 81%, along with the best sensitivity (81%), specificity (81%), precision (81%), F1-score (81%), and G-Mean (0.81). The Random Forest (RF) model also performed well, with an accuracy of 79% and the highest ROC AUC of 0.89, indicating excellent discriminatory ability. The G-Mean scores across all models are summarized in Fig. 3 , illustrating that DT and RF models achieve the best balance between sensitivity and specificity. The ROC curves presented in Fig. 4 , further confirm the superior discriminative performance of the RF and DT models compared to other classifiers. A detailed comparison of model performance across multiple metrics accuracy, sensitivity, specificity, ROC AUC and G-Mean score is shown in Fig. 5 . This figure highlights that DT and RF consistently outperform other models on these key evaluation measures. Finally, confusion matrices for all six classifiers are depicted in a 2×3 grid in Fig. 6 . These matrices provide a granular view of the true positive, true negative, false positive, and false negative predictions, underscoring the strengths of DT and RF in accurately classifying acute respiratory infections. 3.3 SHAP Analysis of Key Predictors The SHAP analysis reveals several important insights regarding the factors influencing the risk of acute respiratory infections (ARI). Notably, the presence of fever (fever_Yes) shows the strongest positive association with ARI, substantially increasing the predicted likelihood of infection. Age also plays a significant role in susceptibility, with specific categories such as 12–23 months, 36–47 months, and 48–59 months demonstrating meaningful variation in risk. Geographic location emerges as another key determinant; divisions including Sylhet, Dhaka, and Chattogram significantly influence the model’s predictions, indicating regional disparities in ARI prevalence. Nutritional status further contributes to risk estimation, as both underweight and obesity (Bmi_Underweight and Bmi_Obesity) are associated with altered susceptibility, underscoring the impact of malnutrition on respiratory health. Additionally, demographic and socioeconomic factors such as sex (male), breastfeeding status (breastfed_Yes), and wealth index (poorest) show moderate importance in the prediction model. Collectively, this comprehensive analysis highlights the multifaceted nature of ARI risk and demonstrates the model’s capacity to integrate clinical, demographic, and environmental variables for accurate disease prediction. These findings are visualized in Fig. 7 , which presents the SHAP summary plot ranking features by their contribution to the model’s predictions. 4. Discussion This study applied various classical machine learning algorithms to investigate acute respiratory tract infections (ARI) and their associated risk factors among children in Bangladesh, with the goal of informing targeted health interventions. Among the models tested, the Decision Tree (DT) algorithm achieved the highest performance, recording 81% accuracy, 81% sensitivity and specificity, 81% precision and F1-score, and an ROC AUC of 88%. The Random Forest (RF) model also performed well, with 79% accuracy and an ROC AUC of 89%, while AdaBoost demonstrated strong predictive capability, attaining 78% accuracy, 78% sensitivity, and an ROC AUC of 86%. Across these models, several key factors were consistently identified as significant predictors of ARI, including child’s age, body mass index (BMI), geographic region (division), household wealth, presence of fever, maternal education, sex of the child, place of residence, recent diarrheal illness, breastfeeding status, child disability, and maternal health challenges. Our results were best with those made in Uganda, which indicated that the random forest model was highly significant for predicting childhood ARI symptoms with an accuracy of 88.70% (24). This could be due to differences in economic status, cultural backgrounds, lifestyles, and specific fields of study. By using SHAP values, the findings revealed that having media exposure, being Fever, Division, Sex, Age, Area, and the normal stunting status were all important variables for the children who had no symptom ARI. Vaccinated status among children was among the sets of predictors studied in Ethiopia, Tigray regional state, and high mortality counters also support this finding (25–27). Effective vaccines in childhood prevent key viral respiratory illnesses (28,29). In the current study, breastfeeding could protect against several acute gastrointestinal and respiratory illnesses. These findings are supported by similar findings in Ethiopia, Cambodia, Uganda, and Kenya (Um. Due to disparities and barriers to health facilities, it is a big problem that people are less likely to seek healthcare (30). Insufficient facilities can hinder mothers from delivering in medical establishments. According to research, children who were born in medical facilities were more likely to visit medical centers for postnatal care and vaccinations, as well as to seek healthcare overall. Mothers may bring their child for medical attention if they experience any ARI symptoms while traveling to these services. Findings are not supported by similar findings (Akinyemi and Morakinyo, 2018; Chilot et al., 2022; Nshimiyimana and Zhou, 2022).Children from rural areas in Sub-Saharan Africa typically get diarrheal diseases as a result of rotavirus infections (33). Compared to children who have never experienced diarrhea, those who have experienced diarrhea within the last two weeks are more likely to experience ARI symptoms. This is consistent with study findings in Ethiopia (25). ARI was significantly more common in children who were small in birth than in children who were average size, where smaller-sized children had an 18.8% higher chance of developing an ARI. This is consistent with study findings in Ethiopia (34), North Jayapura Sub-District (35). Mothers of a child who had access to media exposure were more likely to seek treatment for ARI (36). Mothers are more inclined to pursue medical care when they encounter media that changes their perceptions, beliefs, and societal standards. This also makes mothers more conscious of the significance and urgency of providing healthcare for their children. The impoverished, however, might not be able to afford radio or television. This study was supported in Bangladesh (37). Research has shown that children with stunting are at a higher risk of experiencing acute respiratory infections (ARI). This result is also consistent with research carried out in Ethiopia (38,39). A recent study indicates that malnourished children have compromised immune systems, making them more susceptible to Acute Respiratory Infections (ARI) and various other diseases. There were certain restrictions on this investigation. The key factors concerning acute respiratory tract infections in children under five, as indicated by BDSH data collection, are based on self-reported information, which might have led to some biases in the data. The findings from this study could also contribute to the creation of an online mobile application that predicts acute respiratory tract infections in young children. This tool would allow mothers or other caregivers to recognize early signs of acute tract infections in at-risk children and facilitate access to the necessary treatments. 5. Conclusion This research utilized traditional machine learning approaches to forecast acute respiratory tract infections (ARI) in children within Bangladesh and to pinpoint significant risk factors associated with these infections. Among the models evaluated, Decision Tree, Random Forest, and AdaBoost showed the highest effectiveness, revealing intricate patterns that conventional statistical methods might miss. The identified common risk factors included fever, age, BMI, geographical region, household wealth, maternal education, child’s gender, type of residence, recent history of diarrhea, breastfeeding status, child disabilities, and maternal health issues. These results provide important insights for crafting focused public health initiatives aimed at preventing ARI and enhancing early detection among children. Nonetheless, the study has certain limitations. It depended on self-reported survey information, which could lead to recall or reporting biases, and its cross-sectional design restricts causal interpretations. Future investigations should include more clinical and environmental variables, assess the models in different populations, and look into more sophisticated machine learning methodologies. Creating digital applications based on these models could further assist caregivers and healthcare professionals in identifying ARI risk at an early stage, ultimately leading to better health outcomes for children. Declarations Ethics approval and consent to participate The study used publicly available secondary data from the Bangladesh Demographic and Health Survey (BDHS) 2022. Ethical approval for the BDHS was obtained by the original data collectors from the relevant national ethics review board. Therefore, no additional ethical approval was required for this secondary analysis. The data used were de-identified and publicly available upon request from the DHS Program (https://dhsprogram.com). Consent for publication Not applicable. Competing interests The authors declare that they have no competing interests. Availability of data and materials The datasets used and/or analyzed during the current study are available from the DHS Program upon reasonable request and registration. Data access is granted at: https://dhsprogram.com/data/available-datasets.cfm. Authors’ contributions SKDS (Samrat Kumar Dev Sharma) conceptualized the study, developed the main machine learning analysis, and drafted the manuscript. MYHA (Md. Yusuf Hossain Ador) wrote the main script and contributed to the introduction. JH (Jakir Hossain) paraphrased and refined the introduction. MR (Md. Rokunuzzaman) handled the data processing and management. MK (Md. Kamruzzaman, PhD) reviewed the manuscript critically. MH (Mahmud Hossen) and FC (Futanta Chakma) checked and proofread the final document. All authors read and approved the final manuscript. Acknowledgements The authors would like to thank the Demographic and Health Survey (DHS) Program for providing access to the dataset. Authors’ information All authors are affiliated with the Department of Statistics, Jagannath University, Dhaka-1100, Bangladesh. References Yehuala TZ, Fente BM, Wubante SM, Derseh NM. Exploring machine learning algorithms to predict acute respiratory tract infection and identify its determinants among children under five in Sub-Saharan Africa. Front Pediatr. 2024;12:1388820. Kalayou MH, Kassaw AAK, Shiferaw KB. Empowering child health: Harnessing machine learning to predict acute respiratory infections in Ethiopian under-fives using demographic and health survey insights. BMC Infect Dis. 2024;24(1):338. Kananura RM. Machine learning predictive modelling for identification of predictors of acute respiratory infection and diarrhoea in Uganda’s rural and urban settings. PLOS global public health. 2022;2(5):e0000430. Tadese ZB, Hailu DT, Abebe AW, Kebede SD, Walle AD, Seifu BL, et al. Interpretable prediction of acute respiratory infection disease among under-five children in Ethiopia using ensemble machine learning and Shapley additive explanations (SHAP). Digit Health. 2024;10:20552076241272740. Fenta HM, Zewotir TT, Naidoo S, Naidoo RN, Mwambi H. Factors of acute respiratory infection among under-five children across sub-Saharan African countries using machine learning approaches. Sci Rep. 2024;14(1):15801. Bizzego A, Gabrieli G, Bornstein MH, Deater-Deckard K, Lansford JE, Bradley RH, et al. Predictors of contemporary under-5 child mortality in low-and middle-income countries: A machine learning approach. Int J Environ Res Public Health. 2021;18(3):1315. Islam Pollob SA, Abedin MM, Islam MT, Islam MM, Maniruzzaman M. Predicting risks of low birth weight in Bangladesh with machine learning. PLoS One. 2022;17(5):e0267190. Shaddick G, Thomas ML, Mudu P, Ruggeri G, Gumy S. Half the world’s population are exposed to increasing air pollution. NPJ Clim Atmos Sci. 2020;3(1):23. Gonzales R, Malone DC, Maselli JH, Sande MA. Excessive antibiotic use for acute respiratory infections in the United States. Clinical infectious diseases. 2001;33(6):757–62. Cowley LA, Afrad MH, Rahman SIA, Mamun MM Al, Chin T, Mahmud A, et al. Genomics, social media and mobile phone data enable mapping of SARS-CoV-2 lineages to inform health policy in Bangladesh. Nat Microbiol. 2021;6(10):1271–8. Zhang L, Mendoza-Sassi R, Santos JCH, Lau J. Accuracy of symptoms and signs in predicting hypoxaemia among young children with acute respiratory infection: a meta-analysis. The International journal of tuberculosis and lung disease. 2011;15(3):317–25. De Francisco A, Morris J, Hall AJ, Schellenberg JRMA, Greenwood BM. Risk factors for mortality from acute lower respiratory tract infections in young Gambian children. Int J Epidemiol. 1993;22(6):1174–82. Kassaw AK, Bekele G, Kassaw AK, Yimer A. Prediction of acute respiratory infections using machine learning techniques in Amhara Region, Ethiopia. Sci Rep. 2024;14(1):27968. National Institute of Population Research, (NIPORT) T, of Health M, Welfare F, (ICF) TDHSP. Bangladesh Demographic and Health Survey 2022 [Internet]. Dhaka, Bangladesh; 2022. Available from: https://dhsprogram.com/pubs/pdf/PR148/PR148.pdf Wissmann M, Toutenburg H, others. Role of categorical variables in multicollinearity in the linear regression model. 2007; Chawla N V, Bowyer KW, Hall LO, Kegelmeyer WP. SMOTE: synthetic minority over-sampling technique. Journal of artificial intelligence research. 2002;16:321–57. Marc\’\ilio WE, Eler DM. From explanations to feature selection: assessing SHAP values as feature selection mechanism. In: 2020 33rd SIBGRAPI conference on Graphics, Patterns and Images (SIBGRAPI). 2020. p. 340–7. Liaw A, Wiener M, others. Classification and regression by randomForest. R news. 2002;2(3):18–22. Dietterich TG. Ensemble methods in machine learning. In: International workshop on multiple classifier systems. 2000. p. 1–15. Natras R, Soja B, Schmidt M. Ensemble machine learning of random forest, AdaBoost and XGBoost for vertical total electron content forecasting. Remote Sens (Basel). 2022;14(15):3547. Hosmer Jr DW, Lemeshow S, Sturdivant RX. Applied logistic regression. John Wiley & Sons; 2013. Cortes C, Vapnik V. Support-vector networks. Mach Learn. 1995;20:273–97. Dietterich TG. Ensemble methods in machine learning. In: International workshop on multiple classifier systems. 2000. p. 1–15. Nshimiyimana Y, Zhou Y. Analysis of risk factors associated with acute respiratory infections among under-five children in Uganda. BMC Public Health. 2022;22(1):1209. Tesema GA, Worku MG, Alamneh TS, Teshale AB, Yeshaw Y, Alem AZ, et al. Understanding the rural–urban disparity in acute respiratory infection symptoms among under-five children in Sub-Saharan Africa: a multivariate decomposition analysis. BMC Public Health. 2022;22(1):2013. Gebrerufael GG, Hagos BT. Prevalence and predictors of acute respiratory infection among children under-five years in Tigray regional state, northern Ethiopia: a cross sectional study. BMC Infect Dis. 2023;23(1):743. Mosites EM, Matheson AI, Kern E, Manhart LE, Morris SS, Hawes SE. Care-seeking and appropriate treatment for childhood acute respiratory illness: an analysis of Demographic and Health Survey and Multiple Indicators Cluster Survey datasets for high-mortality countries. BMC Public Health. 2014;14:1–8. Greenberg HB, Piedra PA. Immunization against viral respiratory disease: a review. Pediatr Infect Dis J. 2004;23(11):S254–S261. Zar HJ, Ferkol TW. The global burden of respiratory disease—impact on child health. Vol. 49, Pediatric pulmonology. Wiley Online Library; 2014. p. 430–4. Chilot D, Shitu K, Gela YY, Getnet M, Mulat B, Diress M, et al. Factors associated with healthcare-seeking behavior for symptomatic acute respiratory infection among children in East Africa: a cross-sectional study. BMC Pediatr. 2022;22(1):662. Chilot D, Shitu K, Gela YY, Getnet M, Mulat B, Diress M, et al. Factors associated with healthcare-seeking behavior for symptomatic acute respiratory infection among children in East Africa: a cross-sectional study. BMC Pediatr. 2022;22(1). Akinyemi JO, Morakinyo OM. Household environment and symptoms of childhood acute respiratory tract infections in Nigeria, 2003-2013: A decade of progress and stagnation. BMC Infect Dis. 2018;18(1). Kananura RM. Machine learning predictive modelling for identification of predictors of acute respiratory infection and diarrhoea in Uganda’s rural and urban settings. PLOS global public health. 2022;2(5):e0000430. Anteneh ZA, Hassen HY. Determinants of acute respiratory infection among children in ethiopia: A multilevel analysis from ethiopian demographic and health survey. Int J Gen Med. 2020;13. Mirino R, Dary D, Tampubolon R. Identification of factors causing acute respiratory infection (ARI) of under-fives in community health center work area in north jayapura sub-district. Journal of Tropical Pharmacy and Chemistry. 2022;6(1):15–20. Tazinya AA, Halle-Ekane GE, Mbuagbaw LT, Abanda M, Atashili J, Obama MT. Risk factors for acute respiratory infections in children under five years attending the Bamenda Regional Hospital in Cameroon. BMC Pulm Med. 2018;18:1–8. Chowdhury S, Mishu AA, Rahman MM, Zayed NM. An analysis of factors for acute respiratory infections (ARI) in children under five of age in Bangladesh: a study on DHS, 2014. J Midwifery, Women Heal Gynaecol Nurs. 2020;2(2):26–32. Lema K, Murugan R, Tachbele E, Negussie B. Prevalence and associated factors of pneumonia among under-five children at public hospitals in Jimma zone, South West of Ethiopia, 2018. J Pulmonol Clin Res 2018; 2 (1): 25-31 J Pulmonol Clin Res 2018 Volume 2 Issue. 2018;1. Mandla N, Mackay C, Mda S. Prevalence of severe acute malnutrition and its effect on under-five mortality at a regional hospital in South Africa. South African Journal of Clinical Nutrition. 2022;35(4):149–54. Additional Declarations No competing interests reported. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7081278","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":483043594,"identity":"360c29fc-f6f1-434d-822d-f60421990ecf","order_by":0,"name":"Samrat Kumar Dev Sharma","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA6UlEQVRIiWNgGAWjYFCDAwyMD4AUDx9xyhPAWpgNQFrYSNHCJgFiE9Qi33467cPPHwz2fMd7zCq/5tjJsDEwP3x0A48WgzO5m2f2JDAkzjxzxuy27LZkoMPYjI1z8GlhyN3MwJPAkGBwIy3ttuQ2ZqAWHjZpfFrk+99uZvyTwGBvcP9ZWrHktnrCWhhu5G5mBtrCuOEG8zHGj9sOE9ZicOPtZmaZNAmgX5IPSzNuO87DxkzAL/L9uZsZ39jYAEPsYOPHn9uq7fnZmx8+xuswCADHCAMzD5gkrBwBGH+QonoUjIJRMApGDAAACbpFfeN2WY8AAAAASUVORK5CYII=","orcid":"","institution":"Jagannath University","correspondingAuthor":true,"prefix":"","firstName":"Samrat","middleName":"Kumar Dev","lastName":"Sharma","suffix":""},{"id":483043595,"identity":"f61687ef-dc9f-4f08-b463-7452adad25d2","order_by":1,"name":"Md. Yusuf Hossain Ador","email":"","orcid":"","institution":"Jagannath University","correspondingAuthor":false,"prefix":"","firstName":"Md.","middleName":"Yusuf Hossain","lastName":"Ador","suffix":""},{"id":483043596,"identity":"9984f89d-60a8-430a-8fe7-198f5536cca9","order_by":2,"name":"Md. Rokunuzzaman","email":"","orcid":"","institution":"Jagannath University","correspondingAuthor":false,"prefix":"","firstName":"Md.","middleName":"","lastName":"Rokunuzzaman","suffix":""},{"id":483043597,"identity":"dbdedba4-1e90-40d3-a18b-db1d6710d1ce","order_by":3,"name":"Md. Kamruzzaman","email":"","orcid":"","institution":"Jagannath University","correspondingAuthor":false,"prefix":"","firstName":"Md.","middleName":"","lastName":"Kamruzzaman","suffix":""},{"id":483043598,"identity":"27dae78a-a8a6-4bdd-a699-70dd6a536f64","order_by":4,"name":"Jakir Hossain","email":"","orcid":"","institution":"Jagannath University","correspondingAuthor":false,"prefix":"","firstName":"Jakir","middleName":"","lastName":"Hossain","suffix":""},{"id":483043599,"identity":"d9350d73-7124-4cce-bda6-c818c2e7d5b8","order_by":5,"name":"Mahmud Hossen","email":"","orcid":"","institution":"Jagannath University","correspondingAuthor":false,"prefix":"","firstName":"Mahmud","middleName":"","lastName":"Hossen","suffix":""},{"id":483043600,"identity":"8495ca4b-0de2-46a7-b968-7d8165aa08cf","order_by":6,"name":"Futanta Chakma","email":"","orcid":"","institution":"Jagannath University","correspondingAuthor":false,"prefix":"","firstName":"Futanta","middleName":"","lastName":"Chakma","suffix":""}],"badges":[],"createdAt":"2025-07-09 07:53:35","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7081278/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7081278/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":86473015,"identity":"5af012c1-9444-4294-a79b-d1600a2c7053","added_by":"auto","created_at":"2025-07-11 05:58:58","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":106991,"visible":true,"origin":"","legend":"\u003cp\u003eProposed workflow for predicting acute respiratory infections (ARI) using machine learning. The pipeline includes data preprocessing, feature selection, model training, performance evaluation, and interpretation using SHAP analysis.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-7081278/v1/e8a954d92fcb49c1920fb05b.png"},{"id":86473017,"identity":"0d7244ae-8a7f-4a61-b989-46d69d749804","added_by":"auto","created_at":"2025-07-11 05:58:58","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":78478,"visible":true,"origin":"","legend":"\u003cp\u003eFeature importance ranked by the SHAP values. The plot highlights the top predictors influencing ARI classification.\u003c/p\u003e","description":"","filename":"image2.png","url":"https://assets-eu.researchsquare.com/files/rs-7081278/v1/c2c5d161c7d90de40eceeb9a.png"},{"id":86473417,"identity":"0ce3b38d-a6a7-4d0c-b148-172572b04e0b","added_by":"auto","created_at":"2025-07-11 06:06:58","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":53182,"visible":true,"origin":"","legend":"\u003cp\u003eIllustrates the G-Mean scores for the six machine learning models, highlighting the models’ ability to balance true positive and true negative rates.\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-7081278/v1/89004a4c99a413cdfba4818f.png"},{"id":86473418,"identity":"a7795d1c-6a34-49e8-8167-3d3236721cb2","added_by":"auto","created_at":"2025-07-11 06:06:58","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":139834,"visible":true,"origin":"","legend":"\u003cp\u003eROC curves for the six machine learning models. The Random Forest model achieved the highest area under the curve (AUC), indicating strong classification performance.\u003c/p\u003e","description":"","filename":"image4.png","url":"https://assets-eu.researchsquare.com/files/rs-7081278/v1/233ed1f7d30e75618debb175.png"},{"id":86473981,"identity":"5c95b3e1-d433-412e-ac72-8b2e1145d546","added_by":"auto","created_at":"2025-07-11 06:14:59","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":133390,"visible":true,"origin":"","legend":"\u003cp\u003eComparison of model performance across accuracy, precision, sensitivity, specificity, and F1-score. Decision Tree and Random Forest models performed best overall.\u003c/p\u003e","description":"","filename":"image5.png","url":"https://assets-eu.researchsquare.com/files/rs-7081278/v1/6a129f1c7c4dff237260f98e.png"},{"id":86473021,"identity":"3e8a3e96-b84a-461b-ba74-a46c77e7c4ef","added_by":"auto","created_at":"2025-07-11 05:58:58","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":132082,"visible":true,"origin":"","legend":"\u003cp\u003eConfusion matrices for six machine learning models. Each matrix shows true and false predictions for ARI classification.\u003c/p\u003e","description":"","filename":"6.png","url":"https://assets-eu.researchsquare.com/files/rs-7081278/v1/5a5ac6ec4a2eac601f2d62df.png"},{"id":86473979,"identity":"3c6de895-9216-41f1-adcb-60969f9a20bd","added_by":"auto","created_at":"2025-07-11 06:14:58","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":147369,"visible":true,"origin":"","legend":"\u003cp\u003eSHAP summary plot showing the feature importance for ARI prediction. Features are ranked by their impact on the model's output, with color indicating feature value (red = high, blue = low).\u003c/p\u003e","description":"","filename":"image8.png","url":"https://assets-eu.researchsquare.com/files/rs-7081278/v1/ff3a2eb7970c4f4fff6e33a0.png"},{"id":102185359,"identity":"4733644f-e1c2-4aab-ab91-187dfd087990","added_by":"auto","created_at":"2026-02-09 08:13:19","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1697117,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7081278/v1/2bb53146-8d77-49f4-8f72-7d6bb035fb7d.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Predicting Acute Respiratory Infection Risk in Under–Five Children Using Machine Learning: Evidence from Bangladesh","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eChildren's immune systems are particularly susceptible to illness. Malnutrition, diarrheal illnesses, and acute respiratory infections (ARI) are the leading causes of child sickness and death in developing countries (1). As to the Global Health Observatory's (GHO) 2016 report, the leading causes of death for children under the age of five were preterm birth complications and acute respiratory infections (ARI) (2). ARI can be caused by a variety of pathogens, including viruses, bacteria, and, sometimes, fungi (3). Pneumonia is one of the most prevalent and severe forms of ARI among children under five (4). In 2019, the mortality rate for children younger than five in Sub-Saharan Africa was 78 per 1000 live births, double that of the world and at least 16 times higher than the average of high-income nations (1). The death rate among children under the age of five in Africa was nearly eight times that of Europe. Furthermore, air pollution disproportionately affects children living in low- and middle-income nations (5). In countries with high incomes, breakthroughs in medical care, access to safe sanitation and water, and enhancements in all aspects of life have lowered the infant mortality rate to 5 per one thousand live births (6). Sub-Saharan Africa and Southeast Asia comprised the vast majority of ARI mortality in children under the age of five. ARIs continue to be the most significant burden in developing countries, including Ethiopia and Bangladesh. According to the Bangladesh Demographic and Health Survey (BDHS), ARI causes around 25% of all fatalities among children under the age of five in Bangladesh each year. In accordance to the Bangladesh DHS, 2017-18 - Final Report, 80% of households in Bangladesh cook with solid fuels (coal/lignite, charcoal, wood, straw/shrubs/grass, crops, and animal dung), while 20% use clean fuels (electric power and petroleum-based gas/natural gas/biogas). Furthermore, the primary cause of mortality and morbidity in Bangladesh for diarrhea has been effectively controlled, while the associated risk factors for ARI have been growing by the day (7). Bangladesh is an LMIC, with about 166\u0026nbsp;million people (63%) residing in the countryside (8). Nevertheless, the precise size of ARI, which is currently increasing on a massive scale in Bangladesh, remains unknown. Unlike diarrhea or acute malnutrition, ARI has no acceptable standards, making it impossible to assess case treatment excellence using established criteria. Several research investigations found substantial associations between developing ARI and environmental risk characteristics, such as smoke created by indoor cooking, various types of air pollution from the outdoors, passive smoking, and crowding (9,10). These hazards in children under the age of five produce a variety of severe difficulties, including low birth weight, starvation, measles, pneumonia, and problems with breastfeeding (9). Furthermore, various types of fuel for cooking, poor sanitation conditions, mother education rate, sufficient medication for bowel parasitic organism, place of residence, mother's and child's body mass index (BMI), and financial status index are all possible risk factors for pneumonia/ARI in nations that are developing such as Bangladesh (11,12). Multiple research investigations on healthcare-seeking behavior for signs of ARIs among children under the age of five were undertaken in Bangladesh, Nigeria, East Africa, and rural Kenya (2,13), Ghana, and sub-Saharan Africa (4). Earlier studies used linear and non-linear regression models to investigate the determinants of ARIs in children under the age of five. As long as the researcher is aware, there have been a few previous studies that used algorithmic machine learning to predict ARIs in children younger than five using environmental indicators (5). Previously, conventional regression models were employed to undertake substantial research on the socioeconomic and demographic characteristics related to acute respiratory infections in Ethiopia (13).\u003c/p\u003e\u003cp\u003eRecent developments in machine learning (ML) provide effective methods for analyzing intricate, high-dimensional health data and enhancing disease risk assessment. However, despite their potential, ML models have not been widely applied to evaluate ARI risk in children under five in Bangladesh. The BDHS 2022 dataset presents a chance to implement ML techniques to investigate the interactions between clinical, demographic, and environmental factors.\u003c/p\u003e\u003cp\u003eThis research utilizes various ML classifiers and employs SHapley Additive exPlanations (SHAP) for analyzing feature importance to forecast ARI risk in young children. By evaluating the effectiveness of different models, the study seeks to enhance predictive accuracy and reveal hidden patterns within the data. The results will offer important insights for policymakers and healthcare practitioners to create targeted strategies for the early detection and prevention of ARI, ultimately contributing to a reduction in morbidity and mortality rates among children in Bangladesh.\u003c/p\u003e"},{"header":"2. Methodology","content":"\u003cp\u003eThis study is based on a cross-sectional secondary data analysis using the Bangladesh Demographic and Health Survey (BDHS) 2022, which provides nationally representative data on demographic, reproductive, and health indicators. The BDHS employed a two-stage stratified sampling strategy. Initially, 675 enumeration areas (EAs) were chosen, featuring 438 from rural zones and 237 from urban environments, based on probability proportional to size. In the second stage, 30 households were selected systematically from each EA, resulting in a total sample of 30,078 households (14). The study population was restricted to children under five years of age. Children were included in the analysis if they had symptoms of acute respiratory infection (ARI) within the two weeks preceding the survey, as identified by caregiver-reported short, rapid breathing and or difficult breathing that is chest-related. For the analysis of care-seeking behavior, only children with reported ARI symptoms were included. Care-seeking was defined as seeking advice or treatment from qualified healthcare sources, including public sector providers, private medical facilities, or nongovernmental organizations (NGOs). Children for whom care was sought only from traditional practitioners were excluded from this portion of the analysis, in accordance with BDHS 2022 definitions. After applying these criteria, the final analytic sample consisted of 21,899 children. The overall workflow shows in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\u003ch2\u003e2.1 Study variables\u003c/h2\u003e\u003cp\u003eAcute respiratory infection (ARI) was identifie based on caregiver reports of coughing symptoms occuring within the two weeks prior to the survey. Follwing the BDHS guidelines, the outcome variable was binary, categoried as (0\u0026thinsp;=\u0026thinsp;refer no ARI and 1\u0026thinsp;=\u0026thinsp;refer yes ARI).\u003c/p\u003e\u003cp\u003eA set of independent variables were considered in the analysis based on previous literature review (1,2,4,13). These included age in months (\u0026lt;\u0026thinsp;6, 6\u0026ndash;11, 12\u0026ndash;23, 24\u0026ndash;35, 36\u0026ndash;47, and 48\u0026ndash;59), sex of child (male, female), breastfeeding status (yes or no), recent history of diarrhea and fever (yes or no), and body mass index (BMI), categorized using standard thresholds (underweight, normal, overweight, obese). Socio-demographic variables included place of residence (urban or rural), administrative division (Barisal, Chattogram, Dhaka, Khulna, Mymensingh, Rajshahi, Rangpur, and Sylhet), and household wealth index (poorest, middle, richest). Maternal characteristics included mother\u0026rsquo;s education level (none/pre-primary, primary, secondary, or higher), presence of health difficulties (yes or no), and whether the child had any reported disability (yes or no).\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec4\" class=\"Section2\"\u003e\u003ch2\u003e2.2 Data preprocessing\u003c/h2\u003e\u003cp\u003eThis study, the data preparation process was to ensure the quality and integrity of the dataset prior to analysis. Different key variables including ARI status, age, sex, diarrhea, fever, BMI, area, division, wealth index, mother\u0026rsquo;s education, mother\u0026rsquo;s difficulties had no missing values. However, a substantial proportion of missing data was seen in breastfeeding status (42.22%) and child disability status (37.13%). These missing values were addressed using mode imputation. Categorical variables such as age, sex, breastfeeding status, diarrhea, fever, area, division, child disability, mother\u0026rsquo;s education, mother\u0026rsquo;s health difficulties, wealth index, and BMI category were converted into dummy variables using one-hot encoding. To avoid multicollinearity, the first category of each variable was dropped during encoding (15). The new dummy variables were combined with the dataset, and the original categorical columns were removed. After encoding, the dataset was randomly split into 80% for training and 20% for testing.\u003c/p\u003e\u003cp\u003eTo resolve the issue of class imbalance in the ARI outcome, the Synthetic Minority Over-sampling Technique (SMOTE) was applied. SMOTE generated synthetic instances of the minority class (children with ARI symptoms), resulting in a balanced dataset that improved the stability and generalizability of predictive models (16).\u003c/p\u003e\u003cp\u003eFollowing preprocessing and resampling, feature selection was conducted using a Random Forest model combined with SHAP (SHapley Additive exPlanations) values (17). SHAP values were used to measure how much each feature contributed to the model's prediction. Features with higher SHAP values were considered important and kept for further analysis, while less important ones were removed. This method helped select the most relevant variables and made the model easier to understand and more correct (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eMachine learning algorithms were employed to predict acute respiratory infections (ARI) and identify their contributing factors. Data processing and analysis were carried out in Python, utilizing libraries such as Pandas, Scikit-learn, Imbalanced-learn, NumPy, and Seaborn for tasks including data cleaning, transformation, model training, and evaluation. A predictive model was developed to estimate the risk of ARI and determine the key associated factors.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec5\" class=\"Section2\"\u003e\u003ch2\u003e2.3 Machine Learning Models\u003c/h2\u003e\u003cp\u003eTo predict acute respiratory infections (ARI) and identify associated risk factors, several machine learning algorithms were applied. These models were selected based on their effectiveness in classification tasks, diversity in learning mechanisms, and interpretability.\u003c/p\u003e\u003cp\u003e\u003col\u003e\u003cspan\u003e\u003cli\u003e\u003cp\u003eRandom Forest is a strong ensemble technique that builds several decision trees by employing random samples of the data and features. Final predictions are made through majority voting (for classification) or averaging (for regression). This approach reduces overfitting and improves generalization (Liaw, Wiener and others, 2002).\u003c/p\u003e\u003c/li\u003e\u003c/span\u003e\u003cspan\u003e\u003cli\u003e\u003cp\u003eExtremely Randomized Trees (Extra Trees) operate similarly to Random Forests but introduce greater randomness by selecting split thresholds randomly rather than optimizing them. This can result in faster training and improved generalization performance in certain cases (Dietterich, 2000).\u003c/p\u003e\u003c/li\u003e\u003c/span\u003e\u003cspan\u003e\u003cli\u003e\u003cp\u003eAdaptive Boosting (AdaBoost) is an ensemble technique that combines multiple weak learners, typically shallow decision trees, into a strong classifier. It works iteratively by assigning higher weights to misclassified instances, enabling the model to focus on more difficult cases in subsequent rounds (Natras, Soja and Schmidt, 2022).\u003c/p\u003e\u003c/li\u003e\u003c/span\u003e\u003cspan\u003e\u003cli\u003e\u003cp\u003eLogistic Regression (LR) is a statistical model frequently employed for classifying binary outcomes. It estimates the probability of a class label by applying the logistic function to a linear combination of input features. Logistic regression is appreciated for its simplicity, efficiency, and interpretability (Hosmer Jr, Lemeshow and Sturdivant, 2013).\u003c/p\u003e\u003c/li\u003e\u003c/span\u003e\u003cspan\u003e\u003cli\u003e\u003cp\u003eSupport Vector Classifier (SVC), a form of support vector machine, finds the optimal hyperplane that separates classes in a high-dimensional space. It is particularly effective in handling non-linear relationships through the use of kernel functions (22).\u003c/p\u003e\u003c/li\u003e\u003c/span\u003e\u003cspan\u003e\u003cli\u003e\u003cp\u003eDecision Tree (DT) is a non-parametric model that splits the data into branches based on feature values, forming a tree-like structure. While easy to interpret and visualize, decision trees are susceptible to overfitting, which can be mitigated through pruning or regularization techniques (Dietterich, 2000b).\u003c/p\u003e\u003c/li\u003e\u003c/span\u003e\u003c/ol\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec6\" class=\"Section2\"\u003e\u003ch2\u003e2.4 Evaluation matrices\u003c/h2\u003e\u003cp\u003eModel performance was evaluated using accuracy, sensitivity, specificity, precision, F1- score, G-Mean score and ROC(AUC) score.\u003c/p\u003e\u003cp\u003e\u003cimg src=\"data:image/png;base64,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\" width=\"584\" height=\"341\"\u003e\u003c/p\u003e\u003cp\u003eIn this framework, TP says true positives, TN is negative, FP stands for false positives, and FN denotes false negatives.\u003c/p\u003e\u003c/div\u003e"},{"header":"3. Results","content":"\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e\u003ch2\u003e3.1 Results of the background characteristics\u003c/h2\u003e\u003cp\u003eAmong the 21,899 children under five, ARI prevalence was highest in the 6\u0026ndash;11-month age group (4.3%) and lowest in the 48\u0026ndash;59-month group (1.4%). Male children had a higher ARI rate (2.6%) than females (1.8%). Children who had fever (6.3%) or diarrhoea (3.6%) showed significantly higher ARI rates than those without these symptoms. ARI was slightly more common among urban (2.3%) than rural (2.2%) children. Children with disabilities (4.4%) and those whose mothers reported health difficulties (5.5%) had higher ARI prevalence. Maternal education and household wealth showed minimal variation in ARI rates, which remained around 2.2\u0026ndash;2.4%. Nutritional status was dominated by underweight children (92.2%), with no ARI reported among overweight children. Regionally, the highest ARI rate was found in Mymensingh (6.3%), while Sylhet had the lowest (0.8%) (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e).\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eBackground Characteristics of the Study Populations\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"4\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e\u003cp\u003eVariables\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\" morerows=\"1\" rowspan=\"2\"\u003e\u003cp\u003eFrequency\u003c/p\u003e\u003cp\u003eN (%)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e\u003cp\u003eAcuite Respiratory System (ARI)\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003eNo (%)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003eYes(%)\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003ctr\u003e\u003cth align=\"left\" colspan=\"4\" nameend=\"c4\" namest=\"c1\"\u003e\u003cp\u003e\u003cem\u003eAge (in Month)\u003c/em\u003e\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e0\u0026ndash;5\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e2291 (10.46)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e2212 (96.6)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e79 (3.4)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e6\u0026ndash;11\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e1452 (6.63)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e1390 (95.7)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e62 (4.3)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e12\u0026ndash;23\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e4408 (20.13)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e4295 (97.4)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e113 (2.6)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e24\u0026ndash;35\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e4513 (20.61)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e4411 (97.7)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e102 (2.3)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e36\u0026ndash;47\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e4696 (21.44)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e4626 (98.5)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e70 (1.5)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e48\u0026ndash;59\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e4539 (20.73)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e4476 (98.6)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e63 (1.4)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c4\" namest=\"c1\"\u003e\u003cp\u003e\u003cb\u003eSex (Child)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eFemale\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e10593 (48.37)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e10403 (98.2)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e190 (1.8)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMale\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e11306 (51.63)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e11007 (97.4)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e299 (2.6)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c4\" namest=\"c1\"\u003e\u003cp\u003e\u003cb\u003eBreastfeeding\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eNo\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e566 (2.58)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e553 (97.7)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e13 (2.3)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e21333 (97.42)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e20857 (97.8)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e476 (2.2)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c4\" namest=\"c1\"\u003e\u003cp\u003e\u003cb\u003eDiarrhoea\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eNo\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e20429 (93.29)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e19993 (97.9)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e436 (2.1)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e1470 (6.71)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e1417 (96.4)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e53 (3.6)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c4\" namest=\"c1\"\u003e\u003cp\u003e\u003cb\u003eFever\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eNo\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e16898 (77.16)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e16722 (99.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e176 (1.0)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e5001 (22.84)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e4688 (93.7)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e313 (6.3)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c4\" namest=\"c1\"\u003e\u003cp\u003e\u003cb\u003eArea\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eRural\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e17790 (81.24)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e17396 (97.8)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e394 (2.2)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eUrban\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e4109 (18.76)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e4014 (97.7)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e95 (2.3)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c4\" namest=\"c1\"\u003e\u003cp\u003e\u003cb\u003eChild Disability\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eNo\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e21539 (98.36)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e21066 (97.8)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e473 (2.2)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e360 (1.64)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e344 (95.6)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e16 (4.4)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c4\" namest=\"c1\"\u003e\u003cp\u003e\u003cb\u003eMother Education\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePre-primary\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e2354 (10.75)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e2308 (98.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e46 (2.0)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePrimary\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e5273 (24.08)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e5148 (97.6)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e125 (2.4)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSecondary\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e10830 (49.45)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e10590 (97.8)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e240 (2.2)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eHigher secondary\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e3442 (15.72)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e3364 (97.7)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e78 (2.3)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c4\" namest=\"c1\"\u003e\u003cp\u003e\u003cb\u003eMother Difficulty\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eNo\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e21589 (98.58)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e21117 (97.8)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e472 (2.2)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e310 (1.42)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e293 (94.5)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e17 (5.5)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c4\" namest=\"c1\"\u003e\u003cp\u003e\u003cb\u003eWealth Index\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePoorest\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e5408 (24.70)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e5288 (97.8)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e120 (2.2)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMiddle\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e12823 (58.56)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e12537 (97.8)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e286 (2.2)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eRichest\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e3668 (16.75)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e3585 (97.7)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e83 (2.3)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c4\" namest=\"c1\"\u003e\u003cp\u003e\u003cb\u003eBMI\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eUnderweight\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e20197 (92.23)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e19746 (97.8)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e451 (2.2)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eNormal\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e853 (3.90)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e834 (97.8)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e19 (2.2)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eOverweight\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e817 (3.73)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e32 (100.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0 (0.0)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eObesity\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e32 (0.15)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e798 (97.7)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e19 (2.3)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c4\" namest=\"c1\"\u003e\u003cp\u003e\u003cb\u003eDivision\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eBarishal\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e1953 (8.92)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e1897 (97.1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e56 (2.9)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eChattogram\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e4547 (20.76)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e4457 (98.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e90 (2.0)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eDhaka\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e4299 (19.63)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e4228 (98.3)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e71 (1.7)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eKhulna\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e3023 (13.80)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e2956 (97.8)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e67 (2.2)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMymenshing\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e1293 (5.90)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e1212 (93.7)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e81 (6.3)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eRajshahi\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e2274 (10.38)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e2217 (97.5)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e57 (2.5)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eRangpur\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e2609 (11.91)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e2557 (98.0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e52 (2.0)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSylhet\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e1901 (8.68)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e1886 (99.2)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e15 (0.8)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec9\" class=\"Section2\"\u003e\u003ch2\u003e3.2 Experiments Results\u003c/h2\u003e\u003cp\u003eThe predictive performance of six machine learning classifiers Random Forest (RF), Extra Trees (ET), AdaBoost (AdaB), Logistic Regression (LG), Support Vector Classifier (SVC), and Decision Tree (DT) was evaluated using several metrics including accuracy, sensitivity, specificity, precision, F1-score, ROC AUC, and G-Mean (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e).\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003ePerformance Metrics score for Classical ML Classifiers\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"8\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eModels\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eAccuracy\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003eSensitivity\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003eSpecificity\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e\u003cp\u003ePrecision\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c6\"\u003e\u003cp\u003eF1-score\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c7\"\u003e\u003cp\u003eROC AUC\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c8\"\u003e\u003cp\u003eG-Mean\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eRF\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.79\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.79\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.79\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.79\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.79\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e\u003cb\u003e0.89\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e0.79\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eET\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.76\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.73\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.78\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.77\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.75\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.83\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e0.75\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eAdaB\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.78\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.78\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.79\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.79\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.78\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.86\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e0.78\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eLG\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.75\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.76\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.74\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.75\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.75\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.83\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e0.75\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eSVC\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.74\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.74\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.75\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.74\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.74\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.82\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e0.74\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eDT\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e\u003cb\u003e0.81\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e\u003cb\u003e0.81\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e\u003cb\u003e0.81\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e\u003cb\u003e0.81\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e\u003cb\u003e0.81\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.88\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e\u003cb\u003e0.81\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eAmong the models, the Decision Tree (DT) achieved the highest accuracy of 81%, along with the best sensitivity (81%), specificity (81%), precision (81%), F1-score (81%), and G-Mean (0.81). The Random Forest (RF) model also performed well, with an accuracy of 79% and the highest ROC AUC of 0.89, indicating excellent discriminatory ability.\u003c/p\u003e\u003cp\u003eThe G-Mean scores across all models are summarized in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e, illustrating that DT and RF models achieve the best balance between sensitivity and specificity.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eThe ROC curves presented in Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e, further confirm the superior discriminative performance of the RF and DT models compared to other classifiers.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eA detailed comparison of model performance across multiple metrics accuracy, sensitivity, specificity, ROC AUC and G-Mean score is shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e. This figure highlights that DT and RF consistently outperform other models on these key evaluation measures.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eFinally, confusion matrices for all six classifiers are depicted in a 2\u0026times;3 grid in Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e. These matrices provide a granular view of the true positive, true negative, false positive, and false negative predictions, underscoring the strengths of DT and RF in accurately classifying acute respiratory infections.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e\u003ch2\u003e3.3 SHAP Analysis of Key Predictors\u003c/h2\u003e\u003cp\u003eThe SHAP analysis reveals several important insights regarding the factors influencing the risk of acute respiratory infections (ARI). Notably, the presence of fever (fever_Yes) shows the strongest positive association with ARI, substantially increasing the predicted likelihood of infection. Age also plays a significant role in susceptibility, with specific categories such as 12\u0026ndash;23 months, 36\u0026ndash;47 months, and 48\u0026ndash;59 months demonstrating meaningful variation in risk. Geographic location emerges as another key determinant; divisions including Sylhet, Dhaka, and Chattogram significantly influence the model\u0026rsquo;s predictions, indicating regional disparities in ARI prevalence. Nutritional status further contributes to risk estimation, as both underweight and obesity (Bmi_Underweight and Bmi_Obesity) are associated with altered susceptibility, underscoring the impact of malnutrition on respiratory health. Additionally, demographic and socioeconomic factors such as sex (male), breastfeeding status (breastfed_Yes), and wealth index (poorest) show moderate importance in the prediction model. Collectively, this comprehensive analysis highlights the multifaceted nature of ARI risk and demonstrates the model\u0026rsquo;s capacity to integrate clinical, demographic, and environmental variables for accurate disease prediction.\u003c/p\u003e\u003cp\u003eThese findings are visualized in Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e, which presents the SHAP summary plot ranking features by their contribution to the model\u0026rsquo;s predictions.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e"},{"header":"4. Discussion","content":"\u003cp\u003eThis study applied various classical machine learning algorithms to investigate acute respiratory tract infections (ARI) and their associated risk factors among children in Bangladesh, with the goal of informing targeted health interventions. Among the models tested, the Decision Tree (DT) algorithm achieved the highest performance, recording 81% accuracy, 81% sensitivity and specificity, 81% precision and F1-score, and an ROC AUC of 88%. The Random Forest (RF) model also performed well, with 79% accuracy and an ROC AUC of 89%, while AdaBoost demonstrated strong predictive capability, attaining 78% accuracy, 78% sensitivity, and an ROC AUC of 86%. Across these models, several key factors were consistently identified as significant predictors of ARI, including child\u0026rsquo;s age, body mass index (BMI), geographic region (division), household wealth, presence of fever, maternal education, sex of the child, place of residence, recent diarrheal illness, breastfeeding status, child disability, and maternal health challenges.\u003c/p\u003e\u003cp\u003eOur results were best with those made in Uganda, which indicated that the random forest model was highly significant for predicting childhood ARI symptoms with an accuracy of 88.70% (24). This could be due to differences in economic status, cultural backgrounds, lifestyles, and specific fields of study. By using SHAP values, the findings revealed that having media exposure, being Fever, Division, Sex, Age, Area, and the normal stunting status were all important variables for the children who had no symptom ARI. Vaccinated status among children was among the sets of predictors studied in Ethiopia, Tigray regional state, and high mortality counters also support this finding (25\u0026ndash;27). Effective vaccines in childhood prevent key viral respiratory illnesses (28,29). In the current study, breastfeeding could protect against several acute gastrointestinal and respiratory illnesses. These findings are supported by similar findings in Ethiopia, Cambodia, Uganda, and Kenya (Um. Due to disparities and barriers to health facilities, it is a big problem that people are less likely to seek healthcare (30). Insufficient facilities can hinder mothers from delivering in medical establishments. According to research, children who were born in medical facilities were more likely to visit medical centers for postnatal care and vaccinations, as well as to seek healthcare overall. Mothers may bring their child for medical attention if they experience any ARI symptoms while traveling to these services. Findings are not supported by similar findings (Akinyemi and Morakinyo, 2018; Chilot et al., 2022; Nshimiyimana and Zhou, 2022).Children from rural areas in Sub-Saharan Africa typically get diarrheal diseases as a result of rotavirus infections (33). Compared to children who have never experienced diarrhea, those who have experienced diarrhea within the last two weeks are more likely to experience ARI symptoms. This is consistent with study findings in Ethiopia (25). ARI was significantly more common in children who were small in birth than in children who were average size, where smaller-sized children had an 18.8% higher chance of developing an ARI. This is consistent with study findings in Ethiopia (34), North Jayapura Sub-District (35). Mothers of a child who had access to media exposure were more likely to seek treatment for ARI (36). Mothers are more inclined to pursue medical care when they encounter media that changes their perceptions, beliefs, and societal standards. This also makes mothers more conscious of the significance and urgency of providing healthcare for their children. The impoverished, however, might not be able to afford radio or television. This study was supported in Bangladesh (37). Research has shown that children with stunting are at a higher risk of experiencing acute respiratory infections (ARI). This result is also consistent with research carried out in Ethiopia (38,39). A recent study indicates that malnourished children have compromised immune systems, making them more susceptible to Acute Respiratory Infections (ARI) and various other diseases. There were certain restrictions on this investigation. The key factors concerning acute respiratory tract infections in children under five, as indicated by BDSH data collection, are based on self-reported information, which might have led to some biases in the data. The findings from this study could also contribute to the creation of an online mobile application that predicts acute respiratory tract infections in young children. This tool would allow mothers or other caregivers to recognize early signs of acute tract infections in at-risk children and facilitate access to the necessary treatments.\u003c/p\u003e"},{"header":"5. Conclusion","content":"\u003cp\u003eThis research utilized traditional machine learning approaches to forecast acute respiratory tract infections (ARI) in children within Bangladesh and to pinpoint significant risk factors associated with these infections. Among the models evaluated, Decision Tree, Random Forest, and AdaBoost showed the highest effectiveness, revealing intricate patterns that conventional statistical methods might miss. The identified common risk factors included fever, age, BMI, geographical region, household wealth, maternal education, child\u0026rsquo;s gender, type of residence, recent history of diarrhea, breastfeeding status, child disabilities, and maternal health issues. These results provide important insights for crafting focused public health initiatives aimed at preventing ARI and enhancing early detection among children.\u003c/p\u003e\u003cp\u003eNonetheless, the study has certain limitations. It depended on self-reported survey information, which could lead to recall or reporting biases, and its cross-sectional design restricts causal interpretations. Future investigations should include more clinical and environmental variables, assess the models in different populations, and look into more sophisticated machine learning methodologies. Creating digital applications based on these models could further assist caregivers and healthcare professionals in identifying ARI risk at an early stage, ultimately leading to better health outcomes for children.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe study used publicly available secondary data from the Bangladesh Demographic and Health Survey (BDHS) 2022. Ethical approval for the BDHS was obtained by the original data collectors from the relevant national ethics review board. Therefore, no additional ethical approval was required for this secondary analysis. The data used were de-identified and publicly available upon request from the DHS Program (https://dhsprogram.com).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for publication\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare that they have no competing interests.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAvailability of data and materials\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe datasets used and/or analyzed during the current study are available from the DHS Program upon reasonable request and registration. Data access is granted at: https://dhsprogram.com/data/available-datasets.cfm.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthors\u0026rsquo; contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eSKDS (Samrat Kumar Dev Sharma) conceptualized the study, developed the main machine learning analysis, and drafted the manuscript. MYHA (Md. Yusuf Hossain Ador) wrote the main script and contributed to the introduction. JH (Jakir Hossain) paraphrased and refined the introduction. MR (Md. Rokunuzzaman) handled the data processing and management. MK (Md. Kamruzzaman, PhD) reviewed the manuscript critically. MH (Mahmud Hossen) and FC (Futanta Chakma) checked and proofread the final document. All authors read and approved the final manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors would like to thank the Demographic and Health Survey (DHS) Program for providing access to the dataset.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthors\u0026rsquo; information\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll authors are affiliated with the Department of Statistics, Jagannath University, Dhaka-1100, Bangladesh.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eYehuala TZ, Fente BM, Wubante SM, Derseh NM. Exploring machine learning algorithms to predict acute respiratory tract infection and identify its determinants among children under five in Sub-Saharan Africa. Front Pediatr. 2024;12:1388820. \u003c/li\u003e\n\u003cli\u003eKalayou MH, Kassaw AAK, Shiferaw KB. Empowering child health: Harnessing machine learning to predict acute respiratory infections in Ethiopian under-fives using demographic and health survey insights. BMC Infect Dis. 2024;24(1):338. \u003c/li\u003e\n\u003cli\u003eKananura RM. Machine learning predictive modelling for identification of predictors of acute respiratory infection and diarrhoea in Uganda\u0026rsquo;s rural and urban settings. PLOS global public health. 2022;2(5):e0000430. \u003c/li\u003e\n\u003cli\u003eTadese ZB, Hailu DT, Abebe AW, Kebede SD, Walle AD, Seifu BL, et al. Interpretable prediction of acute respiratory infection disease among under-five children in Ethiopia using ensemble machine learning and Shapley additive explanations (SHAP). Digit Health. 2024;10:20552076241272740. \u003c/li\u003e\n\u003cli\u003eFenta HM, Zewotir TT, Naidoo S, Naidoo RN, Mwambi H. Factors of acute respiratory infection among under-five children across sub-Saharan African countries using machine learning approaches. Sci Rep. 2024;14(1):15801. \u003c/li\u003e\n\u003cli\u003eBizzego A, Gabrieli G, Bornstein MH, Deater-Deckard K, Lansford JE, Bradley RH, et al. Predictors of contemporary under-5 child mortality in low-and middle-income countries: A machine learning approach. Int J Environ Res Public Health. 2021;18(3):1315. \u003c/li\u003e\n\u003cli\u003eIslam Pollob SA, Abedin MM, Islam MT, Islam MM, Maniruzzaman M. Predicting risks of low birth weight in Bangladesh with machine learning. PLoS One. 2022;17(5):e0267190. \u003c/li\u003e\n\u003cli\u003eShaddick G, Thomas ML, Mudu P, Ruggeri G, Gumy S. Half the world\u0026rsquo;s population are exposed to increasing air pollution. NPJ Clim Atmos Sci. 2020;3(1):23. \u003c/li\u003e\n\u003cli\u003eGonzales R, Malone DC, Maselli JH, Sande MA. Excessive antibiotic use for acute respiratory infections in the United States. Clinical infectious diseases. 2001;33(6):757\u0026ndash;62. \u003c/li\u003e\n\u003cli\u003eCowley LA, Afrad MH, Rahman SIA, Mamun MM Al, Chin T, Mahmud A, et al. Genomics, social media and mobile phone data enable mapping of SARS-CoV-2 lineages to inform health policy in Bangladesh. Nat Microbiol. 2021;6(10):1271\u0026ndash;8. \u003c/li\u003e\n\u003cli\u003eZhang L, Mendoza-Sassi R, Santos JCH, Lau J. Accuracy of symptoms and signs in predicting hypoxaemia among young children with acute respiratory infection: a meta-analysis. The International journal of tuberculosis and lung disease. 2011;15(3):317\u0026ndash;25. \u003c/li\u003e\n\u003cli\u003eDe Francisco A, Morris J, Hall AJ, Schellenberg JRMA, Greenwood BM. Risk factors for mortality from acute lower respiratory tract infections in young Gambian children. Int J Epidemiol. 1993;22(6):1174\u0026ndash;82. \u003c/li\u003e\n\u003cli\u003eKassaw AK, Bekele G, Kassaw AK, Yimer A. Prediction of acute respiratory infections using machine learning techniques in Amhara Region, Ethiopia. Sci Rep. 2024;14(1):27968. \u003c/li\u003e\n\u003cli\u003eNational Institute of Population Research, (NIPORT) T, of Health M, Welfare F, (ICF) TDHSP. Bangladesh Demographic and Health Survey 2022 [Internet]. Dhaka, Bangladesh; 2022. Available from: https://dhsprogram.com/pubs/pdf/PR148/PR148.pdf\u003c/li\u003e\n\u003cli\u003eWissmann M, Toutenburg H, others. Role of categorical variables in multicollinearity in the linear regression model. 2007; \u003c/li\u003e\n\u003cli\u003eChawla N V, Bowyer KW, Hall LO, Kegelmeyer WP. SMOTE: synthetic minority over-sampling technique. Journal of artificial intelligence research. 2002;16:321\u0026ndash;57. \u003c/li\u003e\n\u003cli\u003eMarc\\\u0026rsquo;\\ilio WE, Eler DM. From explanations to feature selection: assessing SHAP values as feature selection mechanism. In: 2020 33rd SIBGRAPI conference on Graphics, Patterns and Images (SIBGRAPI). 2020. p. 340\u0026ndash;7. \u003c/li\u003e\n\u003cli\u003eLiaw A, Wiener M, others. Classification and regression by randomForest. R news. 2002;2(3):18\u0026ndash;22. \u003c/li\u003e\n\u003cli\u003eDietterich TG. Ensemble methods in machine learning. In: International workshop on multiple classifier systems. 2000. p. 1\u0026ndash;15. \u003c/li\u003e\n\u003cli\u003eNatras R, Soja B, Schmidt M. Ensemble machine learning of random forest, AdaBoost and XGBoost for vertical total electron content forecasting. Remote Sens (Basel). 2022;14(15):3547. \u003c/li\u003e\n\u003cli\u003eHosmer Jr DW, Lemeshow S, Sturdivant RX. Applied logistic regression. John Wiley \u0026amp; Sons; 2013. \u003c/li\u003e\n\u003cli\u003eCortes C, Vapnik V. Support-vector networks. Mach Learn. 1995;20:273\u0026ndash;97. \u003c/li\u003e\n\u003cli\u003eDietterich TG. Ensemble methods in machine learning. In: International workshop on multiple classifier systems. 2000. p. 1\u0026ndash;15. \u003c/li\u003e\n\u003cli\u003eNshimiyimana Y, Zhou Y. Analysis of risk factors associated with acute respiratory infections among under-five children in Uganda. BMC Public Health. 2022;22(1):1209. \u003c/li\u003e\n\u003cli\u003eTesema GA, Worku MG, Alamneh TS, Teshale AB, Yeshaw Y, Alem AZ, et al. Understanding the rural\u0026ndash;urban disparity in acute respiratory infection symptoms among under-five children in Sub-Saharan Africa: a multivariate decomposition analysis. BMC Public Health. 2022;22(1):2013. \u003c/li\u003e\n\u003cli\u003eGebrerufael GG, Hagos BT. Prevalence and predictors of acute respiratory infection among children under-five years in Tigray regional state, northern Ethiopia: a cross sectional study. BMC Infect Dis. 2023;23(1):743. \u003c/li\u003e\n\u003cli\u003eMosites EM, Matheson AI, Kern E, Manhart LE, Morris SS, Hawes SE. Care-seeking and appropriate treatment for childhood acute respiratory illness: an analysis of Demographic and Health Survey and Multiple Indicators Cluster Survey datasets for high-mortality countries. BMC Public Health. 2014;14:1\u0026ndash;8. \u003c/li\u003e\n\u003cli\u003eGreenberg HB, Piedra PA. Immunization against viral respiratory disease: a review. Pediatr Infect Dis J. 2004;23(11):S254\u0026ndash;S261. \u003c/li\u003e\n\u003cli\u003eZar HJ, Ferkol TW. The global burden of respiratory disease\u0026mdash;impact on child health. Vol. 49, Pediatric pulmonology. Wiley Online Library; 2014. p. 430\u0026ndash;4. \u003c/li\u003e\n\u003cli\u003eChilot D, Shitu K, Gela YY, Getnet M, Mulat B, Diress M, et al. Factors associated with healthcare-seeking behavior for symptomatic acute respiratory infection among children in East Africa: a cross-sectional study. BMC Pediatr. 2022;22(1):662. \u003c/li\u003e\n\u003cli\u003eChilot D, Shitu K, Gela YY, Getnet M, Mulat B, Diress M, et al. Factors associated with healthcare-seeking behavior for symptomatic acute respiratory infection among children in East Africa: a cross-sectional study. BMC Pediatr. 2022;22(1). \u003c/li\u003e\n\u003cli\u003eAkinyemi JO, Morakinyo OM. Household environment and symptoms of childhood acute respiratory tract infections in Nigeria, 2003-2013: A decade of progress and stagnation. BMC Infect Dis. 2018;18(1). \u003c/li\u003e\n\u003cli\u003eKananura RM. Machine learning predictive modelling for identification of predictors of acute respiratory infection and diarrhoea in Uganda\u0026rsquo;s rural and urban settings. PLOS global public health. 2022;2(5):e0000430. \u003c/li\u003e\n\u003cli\u003eAnteneh ZA, Hassen HY. Determinants of acute respiratory infection among children in ethiopia: A multilevel analysis from ethiopian demographic and health survey. Int J Gen Med. 2020;13. \u003c/li\u003e\n\u003cli\u003eMirino R, Dary D, Tampubolon R. Identification of factors causing acute respiratory infection (ARI) of under-fives in community health center work area in north jayapura sub-district. Journal of Tropical Pharmacy and Chemistry. 2022;6(1):15\u0026ndash;20. \u003c/li\u003e\n\u003cli\u003eTazinya AA, Halle-Ekane GE, Mbuagbaw LT, Abanda M, Atashili J, Obama MT. Risk factors for acute respiratory infections in children under five years attending the Bamenda Regional Hospital in Cameroon. BMC Pulm Med. 2018;18:1\u0026ndash;8. \u003c/li\u003e\n\u003cli\u003eChowdhury S, Mishu AA, Rahman MM, Zayed NM. An analysis of factors for acute respiratory infections (ARI) in children under five of age in Bangladesh: a study on DHS, 2014. J Midwifery, Women Heal Gynaecol Nurs. 2020;2(2):26\u0026ndash;32. \u003c/li\u003e\n\u003cli\u003eLema K, Murugan R, Tachbele E, Negussie B. Prevalence and associated factors of pneumonia among under-five children at public hospitals in Jimma zone, South West of Ethiopia, 2018. J Pulmonol Clin Res 2018; 2 (1): 25-31 J Pulmonol Clin Res 2018 Volume 2 Issue. 2018;1. \u003c/li\u003e\n\u003cli\u003eMandla N, Mackay C, Mda S. Prevalence of severe acute malnutrition and its effect on under-five mortality at a regional hospital in South Africa. South African Journal of Clinical Nutrition. 2022;35(4):149\u0026ndash;54. \u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"ARI, children, risk factors, machine learning, fever, geographic division","lastPublishedDoi":"10.21203/rs.3.rs-7081278/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7081278/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eBackground\u003c/h2\u003e\u003cp\u003eChildren\u0026rsquo;s immune systems are particularly vulnerable to infections. In developing countries, malnutrition, diarrheal diseases, and acute respiratory infections (ARI) remain leading causes of illness and mortality among children. This study applied multiple machine learning (ML) techniques to identify key risk factors associated with ARI symptoms in Bangladesh.\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e\u003cp\u003eSecondary data from the Bangladesh Demographic and Health Survey (BDHS) 2022 were analyzed to assess ARI risk factors. Feature selection was conducted using the SHAP algorithm, and six ML classifiers were trained and evaluated with 10-fold cross-validation. Model performance was measured using accuracy, sensitivity, specificity, precision, F1-score, G-mean, and ROC AUC.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e\u003cp\u003eAmong the classifiers, the Decision Tree (DT) model achieved the highest performance across several metrics, with accuracy, sensitivity, specificity, precision, F1-score, and G-mean all at 0.81, and an ROC AUC of 0.88. Random Forest (RF) and AdaBoost (AdaB) also demonstrated strong performance, with RF showing an accuracy of 0.79 and ROC AUC of 0.89, and AdaB achieving an accuracy of 0.78 and ROC AUC of 0.86. Key predictive features included fever, age groups, geographic division, and area of residence.\u003c/p\u003e\u003ch2\u003eConclusions\u003c/h2\u003e\u003cp\u003eThe Decision Tree classifier outperformed other ML models in predicting ARI risk among children under five in Bangladesh, closely followed by Random Forest and AdaBoost. These findings highlight the potential of ML approaches to support targeted interventions by identifying critical risk factors. Government strategies should focus on early detection, improved treatment of fever, and enhanced household conditions to mitigate ARI risk in young children.\u003c/p\u003e","manuscriptTitle":"Predicting Acute Respiratory Infection Risk in Under–Five Children Using Machine Learning: Evidence from Bangladesh","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-07-11 05:58:54","doi":"10.21203/rs.3.rs-7081278/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"92206524-06b6-40f2-b648-dcb621006d8a","owner":[],"postedDate":"July 11th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2026-02-09T08:10:23+00:00","versionOfRecord":[],"versionCreatedAt":"2025-07-11 05:58:54","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7081278","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7081278","identity":"rs-7081278","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}