Advancing Smart Transportation via AI for Sustainable Traffic Solutions in Saudi Arabia

Advancing Smart Transportation via AI for Sustainable Traffic Solutions in Saudi Arabia | 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 Advancing Smart Transportation via AI for Sustainable Traffic Solutions in Saudi Arabia GOPICHAND BANDARUPALLI This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-5389235/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 The Saudi Arabian government has committed more than $100 billion (USD) to improving the country's transportation infrastructure, in line with Vision 2030 and the Sustainable Development Goals (SDGs) of the United Nations. The National Center for Transportation Safety (NCTS), which focuses on road safety, and the "Rental Contracts" initiative are two examples of the infrastructure development projects for which the FY2022 budget allotted 42 billion SAR. On the other hand, as cities become more populated, traffic congestion has worsened, making living more difficult. In response to these issues, the government is putting in place intelligent transportation systems that use Artificial Intelligence (AI) methods to predict traffic patterns and provide drivers with other routes that cut down on travel time. These AI-driven forecasts are anticipated to lessen traffic-related problems like pollution and health hazards, supporting the country's larger objectives for sustainable infrastructure. AI models, such as Random Forest (RF), Gated Recurrent Units (GRU), and Long Short-Term Memory (LSTM), have been shown to be useful in traffic prediction based on empirical results. With a Mean Square Error (MSE) of 10.5, a Root Mean Square Error (RMSE) of 3.24, a Mean Absolute Error (MAE) of 2.15, and a Mean Absolute Percentage Error (MAPE) of 3.9%, the LSTM model outperformed both the RF and GRU models. These findings demonstrate how AI-driven models may help Saudi Arabia create transportation systems that are reliable, effective, and sustainable. Artificial Intelligence and Machine Learning Artificial Intelligence Deep Learning Support Vector Machines Traffic Congestion United Nations' Figures Figure 1 Figure 2 Figure 3 Figure 4 I. INTRODUCTION Saudi Arabia's Vision 2030 focuses on self-driving cars, which shows how important changing transportation is. To support this vision, a smart transportation system needs to be set up quickly. This system can help deal with traffic jams effectively. During Hajj and Umrah seasons, millions of people visit the Holy cities and other major towns. This causes traffic jams in those areas and nearby places. These traffic problems are in line with the Kingdom's Sustainable Development Goals (SDGs). The main aim of these goals is to improve people's lives by managing movement better. This smart transportation system will predict where traffic jams might happen. This will help to regulate traffic better and reduce accidents, injuries, and the time it takes to travel. It will make life better for everyone. Also, when there is a lot of traffic, more energy gets used and more pollution happens. So, having a system that predicts traffic jams is more important than ever. Artificial Intelligence (AI) methods will be used to create a strong transportation model that can predict where congestion might happen accurately [1], [2], [3], [4], [5], [6]. This system should be able to detect unexpected events like roadblocks or accidents. It can then control them well and keep traffic moving smoothly, especially during busy times. More research will be done to make sure the predictions are accurate, and the system works efficiently. This will help reduce traffic jams and their problems. The whole idea supports the country's Sustainable Development Goals (SDGs), which also focus on having strong infrastructure. This sets the stage for a transportation network that is good for the environment and makes everything move faster because all the parts work well together. The study is structured into five main sections. Firstly, it provides background information to set the context for the research. Secondly, it reviews related works to highlight existing knowledge and gaps in the field. Thirdly, it describes the materials and methods used in the study, detailing the approach taken to gather and analyse data. Fourthly, it presents the results of the experimental analysis, showcasing the findings and any significant observations made during the research process. Finally, it concludes the study by summarizing key findings and offering suggestions for future research directions and areas of improvement in the field. II. BACKGROUND Around the world, many cities have a big problem with traffic. This means roads often have too many cars causing them to move slowly or stop a lot especially when lots of people are going to or from work. There are many reasons for this like more people moving to cities, more people being born, also the economy growing and not enough good roads and public transport. This problem isn't just annoying; it can cause protests, make businesses lose money, and harm nature. It doesn't only affect the city where it happens, but also nearby places. It touches every part of life in a society. Traffic jams in streets, busy city centers, or crowded road crossings are all signs of traffic congestion. It happens when there are too many people trying to use the roads and there are not enough ways for them to travel easily. This happens because of things like where people live and work, how many cars there are, how easy it is to use buses and trains and how the cities are designed. This problem affects healthcare, protecting the environment, and making the economy better. Spending a long time in traffic and using more fuel makes the air dirty and loud. This makes life harder for poor people and makes it tough for them to get jobs, healthcare, and education. Traffic jams also make these businesses lose money because they cannot work well, waste fuel and make it cost more to move things around. To fix this, cities need to plan better for transport in a way that helps everyone and does not hurt the environment. Many cities worldwide are trying different ways to solve this problem of too many cars on the roads. Several cities are also spending a lot of money on mass transit systems. Others are building more sidewalks for people to walk on. This is to make it harder for people to use their own cars which can cause traffic jams, especially during busy times like rush hours in downtown areas. Another idea is for governments to charge money for driving on certain roads at certain times. This is called congestion charging. It is signified to make people think twice about driving during busy times. Some cities are also using smart technology to control traffic lights which change depending on how many cars are on the road. They can help cars avoid getting stuck in traffic jams. Technology is also helping these cities to manage traffic better. With big data and live monitoring, cities can see what is happening on the roads in real-time. This helps them make decisions to keep traffic moving. It also helps them respond quickly if something goes wrong, like a car breaking down. All these efforts aim to make traffic flow better and reduce the chances of accidents. By encouraging people to use public transportation, walk, or carpool, cities hope to ease congestion and make streets safer for everyone. They also want to make sure that important areas, like hospitals and schools, stay accessible even when there's a lot of traffic. III. RELATED WORKS Model-driven prediction approaches have been extensively utilized in the examination and projection of traffic flow dynamics. These methodologies incorporate various techniques such as moving averages [7], exponential smoothing [8], the Kalman filter [9], and Autoregressive Moving Averages (ARIMA) models [10]. For instance, [11] implemented the exponential smoothing technique to analyze and forecast traffic flow within urban transportation networks. [12] built upon this groundwork by improving a spatio-temporal ARIMA model, customizing it for the accurate forecasting of traffic flow in urban environments at five-minute intervals. Similarly, [13] merged the ARIMA model with Support Vector Regression (SVR), thus enhancing the precision of traffic prediction efforts. Additionally, [14] introduced a non-parametric K-Nearest Neighbor (KNN) model specifically tailored for predicting urban road traffic speed and flow patterns. Finally, [15] developed a dynamic multi-interval traffic flow prediction approach based on KNN non-parametric regression models. These research endeavors collectively highlight the effectiveness and adaptability of model-driven prediction strategies in improving our comprehension and forecasting capabilities within the domain of traffic flow analysis. However, in comparison to model-driven methodologies, data-driven strategies employing Artificial Neural Network (ANN) have demonstrated their effectiveness in predicting traffic flow such as [16], [17]. [18] proposed a traffic flow prediction model based on Multi-Layer Perceptron (MLP) neural networks. [19] employed an ANN to anticipate short-term traffic flow, utilizing factors like traffic volume, speed, density, and time. [20] utilized DL methods to capture complex traffic flow patterns without prior explicit knowledge. Following this, [21] introduced a technique utilizing Convolutional Neural Networks (CNNs) for estimating traffic flow speed across extensive networks. They verified the efficacy of CNNs in extracting localized features through convolutional filter layers, thus mimicking spatial dependencies within urban areas using sliding windows. [17] devised a comprehensive DL framework merging residual and convolutional networks. This framework integrated traffic flow data and weather conditions as inputs, facilitating the capture of temporal similarities across various road segments and enhancing traffic flow prediction accuracy. Additionally, [22] introduced a hybrid DL approach for short-term traffic flow prediction. Furthermore, [23] introduced Long Short-Term Memory (LSTM) neural networks for estimating traffic flow within a traffic network's domain. Recent studies have also explored various methods to analyze and predict commuter travel patterns. For instance, [24] utilized LSTM networks to estimate commuter travel time, while [10] integrated linear regression with LSTM for forecasting travel demand in urban settings like New York. However, these methods often overlook the combined spatial and temporal aspects of city-scale traffic flow, which are crucial for precise predictions. CNN are adept at capturing local spatial features, whereas LSTM networks excel in capturing long-term dependencies in sequential data [25], [26], [27], [28], [29], [30], [31]. The COVID-19 pandemic has significantly influenced travel behavior, prompting innovative applications of CNN and LSTM networks. Moreover, [32] employed this hybrid model to estimate traffic flow during rainfall events. Building upon these insights, our research introduces a tailored data-driven model to analyze the spatial and temporal aspects of city-scale traffic flow, with a specific focus on short-term predictions. A summary of the studies listed above is presented in Table I. TABLE I COMMON FEATURES IN TRAFFIC FLOW PREDICTION STUDIES Study Time Series Data Spatial Data Temporal Data Short-Term Prediction Long-Term Prediction Hybrid Model Non-Parametric Model Weather Consideration COVID-19 Impact [7] ✔ ❌ ✔ ✔ ❌ ❌ ❌ ❌ ❌ [8] ✔ ❌ ✔ ✔ ❌ ❌ ❌ ❌ ❌ [9] ✔ ❌ ✔ ✔ ❌ ❌ ❌ ❌ ❌ [10] ✔ ❌ ✔ ✔ ✔ ❌ ❌ ❌ ❌ [11] ✔ ❌ ✔ ✔ ❌ ❌ ❌ ❌ ❌ [12] ✔ ✔ ✔ ✔ ❌ ✔ ❌ ❌ ❌ [13] ✔ ✔ ✔ ✔ ❌ ✔ ❌ ❌ ❌ [14] ✔ ✔ ✔ ✔ ❌ ❌ ✔ ❌ ❌ [15] ✔ ✔ ✔ ✔ ✔ ✔ ✔ ❌ ❌ [16] ✔ ✔ ✔ ✔ ❌ ❌ ❌ ❌ ❌ [17] ✔ ✔ ✔ ✔ ✔ ✔ ❌ ✔ ❌ [18] ✔ ✔ ✔ ✔ ❌ ❌ ❌ ❌ ❌ [19] ✔ ✔ ✔ ✔ ❌ ❌ ❌ ❌ ❌ [20] ✔ ✔ ✔ ✔ ✔ ✔ ❌ ❌ ❌ [21] ✔ ✔ ✔ ✔ ❌ ✔ ❌ ❌ ❌ [22] ✔ ✔ ✔ ✔ ✔ ✔ ❌ ❌ ❌ [23] ✔ ✔ ✔ ✔ ✔ ✔ ❌ ❌ ❌ [24] ✔ ❌ ✔ ✔ ❌ ❌ ❌ ❌ ❌ [32] ✔ ✔ ✔ ✔ ✔ ✔ ❌ ✔ ❌ IV. MATERIALS AND METHODS A. Dataset Analysis Urban areas are complex systems when it comes to traffic, and they change dramatically throughout different days and times according to data analysis. What we call weekdays have extremely high volumes of traffic in specific time brackets usually called rush hours. This period often starts at 6 am to 8 am and from 4 pm to 6 pm in the evening whereby most people are either going or coming from work thereby heavily impacting on congestion levels. Fridays show unique deviation from what would otherwise be anticipated as typical weekday traffic patterns; despite being considered part of weekdays, congestion is not experienced during normal working hours but there is still increased road usage though not as much as other days of the week due to leisure activities and social gatherings. Monday is the only day that commuters commute consistently throughout these days because it is their first day back at work after taking a weekend off. Therefore, temporal factors should not be disregarded when designing effective management systems since they offer important insights into the potential intensity of bottlenecks at specific times if nothing is done to alleviate them. The dataset description used in this study is compiled in Table II. TABLE II SUMMARY OF TEMPORAL TRAFFIC DATASET Day Time Period Traffic Volume Key Observations Weekdays 06:00–08:00 High Morning rush hour due to work commutes 16:00–18:00 High Evening rush hour due to work commutes Friday 06:00–08:00 Lower Reduced morning congestion compared to other weekdays 16:00–18:00 Moderate Evening traffic due to social and recreational activities Weekends Various Variable Much lesser uniformed traffic patterns contrast to weekdays B. Model Analysis Temporal traffic pattern insights are incredibly important to urban planners and decision-makers in their effort to reduce congestion and plan for infrastructural growth as well as transport system development towards building resilient cities that can withstand shocks such as earthquakes, among others. There were three algorithms used for predicting future transportation state which are Random Forests (RF), Gated Recurrent Units (GRUs) and Long Short-Term Memories (LSTMs). Among the performance metrics used to evaluate the RF model with 100 estimators included Mean Squared Error (MSE), Root Mean Square Error (RMSE), Mean Absolute Error(MAE), Mean Absolute Percentage Error(MAPE) ,Error cost, Outlier sensitivity, Model complexity etc then compared forecast against actual data for visualizing how good was our model? Like this, the GRU model, which consisted of a single GRU layer followed by two dense layers, underwent extensive training and evaluation. Performance measures were calculated, and predictions were then compared to the real data. However, a type error occurred during the visualization step because the test set, and predictions had different formats. This issue was resolved through transforming the forecasts into a numpy array. The training, assessment, and presentation procedures for the LSTM model—which consists of one LSTM layer followed by two dense layers—should be mentioned, among other things. It is also important to note that Fig. 1 displays the graphical representation errors distribution for each model, providing some insight into the distribution and concentration sections where such errors in prediction cluster. V. EXPERIMENTAL ANALYSIS In this section, we discuss traffic forecast techniques using three models: RF, GRU, and LSTM. To enable the chosen models to use the dataset, we first do considerable pre-processing on it. These procedures include encoding categorical information like the day of the week and traffic conditions, as well as standardizing time to a 24-hour format. The data set is split into training and testing sets for model training and evaluation, accordingly, following the pre. Our investigation begins with the RF model, which is well known for its ability to handle complex information. The RF model offers insights into its predicted performance through an extensive evaluation process. Specifically, its ensemble learning approach exhibits competitive performance metrics, indicating strong performance in tasks such as traffic prediction where multiple factors interact to influence the result. The metrics that were employed by the RF model are summarized in Table III and Fig. 2 . The mean square error (MSE) for the RF model between the actual and projected values is 15.6. RMSE serves as a valuable metric for assessing the magnitude of error in predictions. It is an approach to understand how far off predictions are from real values. RMSE is calculated by taking the square root of MSE, which stands for Mean Square Error. When RMSE is low, the predictions are generally close to the actual values. So, the relatively low MSE indicates a level of reliability in the model's predictions. For example, if a model has a mean absolute error of 2.75, it means that, on average, the predictions are off by about 2.75 units. This shows that the model makes modest mistakes in its predictions, which is a good sign for its reliability. Another measure that can be looked at is MAPE or Mean Absolute Percentage Error. A MAPE of 5.3% means that, on average, the predictions are within 5.3% of the actual values. This is an important strategy in traffic forecasting because it tells us that the model's predictions are generally pretty close to reality. In real-world traffic forecasting, it is often hard to get detailed or reliable historical data to base predictions on. So, having a model that can make accurate predictions even with such limited data is a crucial role. While managing traffic in cities, it is okay to make some mistakes in predictions because they usually don't cause a big problem. Mistakes might mean a slight delay in traffic plans, but they're not usually too serious. The costs of these mistakes are manageable and can be dealt with by taking suitable actions at strategic points along important paths in the city. One of the reasons why the RF model, or Random Forest model, is good for traffic forecasting is because they are not easily affected by the outliers. These Outliers are unusual or abnormal data points that can sometimes throw off predictions. But the RF model is designed to handle these kinds of irregularities. It is like having a traffic manager who can deal with unexpected events like accidents or sudden increases in traffic volume even without getting confused. This means that the model can make accurate predictions even while facing random data points. The RF model works by building many decision trees and then combining them together. This makes them really good at understanding complex relationships between different factors that affect traffic. It is like having a team of experts who know all the ins and outs of the traffic patterns. However, because the model is so complex, it needs a lot of processing power to work effectively. Overall, the RF model is effective in traffic prediction because it's accurate and reliable. It can handle moderate errors without causing significant problems, making it suitable for real-world scenarios where some errors are expected. It is not easily affected by outliers; it can make accurate predictions even when faced with abnormal data points. So, if there's a need for a model that can make accurate predictions for traffic forecasting, the RF model is a good option to consider. This is especially true if the data points are frequently found along major highways with numerous entrances and exits close to one another over short distances, heavy traffic during peak hours, and sharp changes over time due to various factors like accidents, road works, etc. TABLE III RANDOM FOREST MODEL PERFORMANCE METRICS Metric Value Mean Squared Error (MSE) 15.6 Root Mean Squared Error (RMSE) 3.95 Mean Absolute Error (MAE) 2.75 Mean Absolute Percentage Error (MAPE) 5.3% Error Cost Moderate Outlier Sensitivity Low Model Complexity High Next, our focus shifts to Recurrent Neural Network (RNN) architectures, beginning with the GRU model. By leveraging its ability to capture sequential dependencies in data, the GRU model is trained and assessed. Despite showing promising results, with significant improvements over traditional ML methods, it does not achieve optimal performance metrics compared to the RF model. Table IV outlines the performance characteristics of the GRU model and provides valuable comparisons with the RF model previously evaluated as depicted in Fig. 3 . The mean squared variance between the expected and actual values compared to RF is less than the MSE of 12.8, which indicates that the GRU model can capture temporal correlations in traffic data. The model's RMSE of 3.58, which places it higher in terms of overall predictive performance than the RF model, further demonstrates its capacity to foresee with a smaller margin of inaccuracy. The mean absolute error (MAE) of the GRU model is 2.45, which is a lower value and highlights the model's accuracy in predicting traffic patterns. Furthermore, with a MAPE of 4.7% indicating that it is within 4.7% of the real values, the GRU model performs somewhat better than the RF model. Despite these promising metrics, the GRU model has moderate error costs, similar to the RF model, indicating that even though it makes generally accurate predictions, prediction errors can still have adverse operational and economic consequences that must be managed in real-world applications. While handling irregular data points better than many traditional ML models, the GRU model is not as robust as the RF model due to its moderate sensitivity to outliers. The model's performance in scenarios where anomalies occur frequently, such as traffic accidents or sudden increases in volume, may be affected by this moderate sensitivity. In terms of model complexity, the GRU is classified as medium. Due to its gating mechanisms and sequential nature, it is more complex by nature than typical ML models, but not as complex as the ensemble-based RF model. This medium complexity is a suitable option for capturing temporal trends without unduly straining computational resources as it balances computational needs and predictive capabilities. TABLE IV GRU MODEL PERFORMANCE METRICS Metric Value Mean Squared Error (MSE) 12.8 Root Mean Squared Error (RMSE) 3.58 Mean Absolute Error (MAE) 2.45 Mean Absolute Percentage Error (MAPE) 4.7% Error Cost Moderate Outlier Sensitivity Moderate Model Complexity Medium Next, we introduced the LSTM model that has been known for a long time as the best in long-term dependency recognition in sequence data. In this light, traffic prediction tasks are conducted to see how well it can perform. However, there should be some more tests against an RF model, so we know what works better. Known for its depth and memory cells with specialized functions, the LSTM model has proved effective in traffic prediction. The training phase had several other models whose performance metrics were not as good as those of this one since during evaluation it achieved the lowest test loss ever recorded among all considered models. This demonstrates that the only algorithm capable of making such complex predictions about traffic patterns would have been the Long Short-Term Memory (LSTM) model, which processes inputs over time steps into outputs across variable sequence lengths until convergence on some fixed-point. What more could you ask for from a Long Short-Term Memory (LSTM) model? Furthermore, the results obtained from examining the plots depicted in Table V and Fig. 4 provide us with indications regarding the potential success or failure of a certain predictive capability in comparison to other similar ones, such as the two displayed here, where they differ. The difference between the two models' accuracy in predicting all points examined so far throughout our research into each model's strengths and weaknesses is ΔY (Actual – Predicted), which is always within ΔX rather than zero. This indicates that both models perform equally poorly in predicting weak areas closer to either end point, possibly in part as none of them recognize features outside a certain range of values. As an illustration of the lack of consistency between anticipated forecasts made based solely on this type of proof, we can see that, between various points along the x-axis, the most faraway ones are more closely related than the two most adjacent indicated values themselves farther apart, but never precisely identical distance away from each other. This still fails to sufficiently account for the least squares fits noticed. TABLE V LSTM MODEL PERFORMANCE METRICS Metric Value Mean Squared Error (MSE) 10.5 Root Mean Squared Error (RMSE) 3.24 Mean Absolute Error (MAE) 2.15 Mean Absolute Percentage Error (MAPE) 3.9% Error Cost Low Outlier Sensitivity Low Model Complexity High The experimental investigation showed that a variety of traffic forecast models, ranging from more sophisticated deep learning (DL) structures like Gated Recurrent Units (GRU) and Long Short-Term Memory (LSTM) to more conventional techniques like Random Forest (RF), are effective. Each model in the realm of traffic prediction has demonstrated pros and cons of its own. However, the LSTM model outperformed the others, achieving the highest accurate rate in traffic trend predicting. These results are critical to transportation planning and management because they offer practical guidance on enhancing system efficiency and traffic flow optimization. VI. CONCLUSION AND FUTURE WORKS In general, our test indicates the many traffic prediction methods that can be used. The study shows that both regular machine learning and deep learning are good at predicting things. We looked at RF, GRU, and LSTM models and found them to be strong. Among them, the LSTM model looks especially promising. It's good at understanding long-term connections between traffic data and complicated patterns. These findings matter a lot for planning and managing transportation. They help make traffic flow better and systems more efficient. In the future, we should focus on making DL models like LSTM even better. We can also mix them with other methods to make them stronger. This indicates that using real-time data to make predictions is more accurate and we need to make sure these predictions work well in real life, especially in big cities with lots of traffic. Additionally, they will investigate how things like weather and the city changes affect traffic. This will help us to make even better models that can handle the challenges of city life. Declarations A. Funding: No funds, grants, or other support was received. B. Conflict of Interest: The authors declare that they have no known competing for financial interests or personal relationships that could have appeared to influence the work reported in this paper. C. Data Availability: Data will be made on reasonable request. D. Code Availability: Code will be made on reasonable request. References H. Habib, G. S. Kashyap, N. Tabassum, and T. Nafis, “Stock Price Prediction Using Artificial Intelligence Based on LSTM– Deep Learning Model,” in Artificial Intelligence & Blockchain in Cyber Physical Systems: Technologies & Applications , CRC Press, 2023, pp. 93–99. doi: 10.1201/9781003190301-6. S. Wazir, G. S. Kashyap, K. Malik, and A. E. I. Brownlee, “Predicting the Infection Level of COVID-19 Virus Using Normal Distribution-Based Approximation Model and PSO,” Springer, Cham, 2023, pp. 75–91. doi: 10.1007/978-3-031-33183-1_5. G. S. Kashyap, D. Mahajan, O. C. Phukan, A. Kumar, A. E. I. Brownlee, and J. Gao, “From Simulations to Reality: Enhancing Multi-Robot Exploration for Urban Search and Rescue,” Nov. 2023, Accessed: Dec. 03, 2023. [Online]. Available: https://arxiv.org/abs/2311.16958v1 G. S. Kashyap, A. Siddiqui, R. Siddiqui, K. Malik, S. Wazir, and A. E. I. Brownlee, “Prediction of Suicidal Risk Using Machine Learning Models,” Dec. 25, 2021. Accessed: Feb. 04, 2024. [Online]. Available: https://papers.ssrn.com/abstract=4709789 G. S. Kashyap, K. Malik, S. Wazir, and R. Khan, “Using Machine Learning to Quantify the Multimedia Risk Due to Fuzzing,” Multimed. Tools Appl. , vol. 81, no. 25, pp. 36685–36698, Oct. 2022, doi: 10.1007/s11042-021-11558-9. G. S. Kashyap et al. , “Detection of a facemask in real-time using deep learning methods: Prevention of Covid 19,” Jan. 2024, Accessed: Feb. 04, 2024. [Online]. Available: https://arxiv.org/abs/2401.15675v1 Q. Liu, E. Chung, and L. Zhai, “Fusing moving average model and stationary wavelet decomposition for automatic incident detection: case study of Tokyo Expressway,” J. Traffic Transp. Eng. (English Ed. , vol. 1, no. 6, pp. 404–414, Dec. 2014, doi: 10.1016/S2095-7564(15)30290-7. K. Y. Chan, T. S. Dillon, J. Singh, and E. Chang, “Traffic flow forecasting neural networks based on exponential smoothing method,” in Proceedings of the 2011 6th IEEE Conference on Industrial Electronics and Applications, ICIEA 2011 , 2011, pp. 376–381. doi: 10.1109/ICIEA.2011.5975612. S. V. Kumar, “Traffic Flow Prediction using Kalman Filtering Technique,” in Procedia Engineering , No longer published by Elsevier, Jan. 2017, pp. 582–587. doi: 10.1016/j.proeng.2017.04.417. T. Mai, B. Ghosh, and S. Wilson, “Short-term traffic-flow forecasting with auto-regressive moving average models,” Proc. Inst. Civ. Eng. Transp. , vol. 167, no. 4, pp. 232–239, May 2014, doi: 10.1680/tran.12.00012. B. M. Williams, P. K. Durvasula, and D. E. Brown, “Urban freeway traffic flow prediction: Application of seasonal autoregressive integrated moving average and exponential smoothing models,” Transp. Res. Rec. , no. 1644, pp. 132–141, Jan. 1998, doi: 10.3141/1644-14. Q. Ding, X. Wang, X. Zhang, and Z. Sun, “Forecasting traffic volume with space-time ARIMA model,” in Advanced Materials Research , Trans Tech Publications Ltd, 2011, pp. 979–983. doi: 10.4028/www.scientific.net/AMR.156-157.979. L. Li, S. He, J. Zhang, and B. Ran, “Short-term highway traffic flow prediction based on a hybrid strategy considering temporal–spatial information,” J. Adv. Transp. , vol. 50, no. 8, pp. 2029–2040, Dec. 2016, doi: 10.1002/atr.1443. D. Xia, B. Wang, H. Li, Y. Li, and Z. Zhang, “A distributed spatial-temporal weighted model on MapReduce for short-term traffic flow forecasting,” Neurocomputing , vol. 179, pp. 246–263, Feb. 2016, doi: 10.1016/j.neucom.2015.12.013. H. Chang, Y. Lee, B. Yoon, and S. Baek, “Dynamic near-term traffic flow prediction: System-oriented approach based on past experiences,” IET Intell. Transp. Syst. , vol. 6, no. 3, pp. 292–305, Sep. 2012, doi: 10.1049/iet-its.2011.0123. T. Kim, S. Sharda, X. Zhou, and R. M. Pendyala, “A stepwise interpretable machine learning framework using linear regression (LR) and long short-term memory (LSTM): City-wide demand-side prediction of yellow taxi and for-hire vehicle (FHV) service,” Transp. Res. Part C Emerg. Technol. , vol. 120, p. 102786, Nov. 2020, doi: 10.1016/j.trc.2020.102786. J. Zhang, Y. Zheng, and D. Qi, “Deep spatio-temporal residual networks for citywide crowd flows prediction,” in 31st AAAI Conference on Artificial Intelligence, AAAI 2017 , AAAI press, Feb. 2017, pp. 1655–1661. doi: 10.1609/aaai.v31i1.10735. Z. Zhu, B. Peng, C. Xiong, and L. Zhang, “Short-term traffic flow prediction with linear conditional Gaussian Bayesian network,” J. Adv. Transp. , vol. 50, no. 6, pp. 1111–1123, Oct. 2016, doi: 10.1002/atr.1392. K. Kumar, M. Parida, and V. K. Katiyar, “Short Term Traffic Flow Prediction for a Non Urban Highway Using Artificial Neural Network,” Procedia - Soc. Behav. Sci. , vol. 104, pp. 755–764, Dec. 2013, doi: 10.1016/j.sbspro.2013.11.170. Y. Lv, Y. Duan, W. Kang, Z. Li, and F. Y. Wang, “Traffic Flow Prediction with Big Data: A Deep Learning Approach,” IEEE Trans. Intell. Transp. Syst. , vol. 16, no. 2, pp. 865–873, Apr. 2015, doi: 10.1109/TITS.2014.2345663. X. Ma, Z. Tao, Y. Wang, H. Yu, and Y. Wang, “Long short-term memory neural network for traffic speed prediction using remote microwave sensor data,” Transp. Res. Part C Emerg. Technol. , vol. 54, pp. 187–197, May 2015, doi: 10.1016/j.trc.2015.03.014. Y. Wu and H. Tan, “Short-term traffic flow forecasting with spatial-temporal correlation in a hybrid deep learning framework,” Dec. 2016, Accessed: May 22, 2024. [Online]. Available: https://arxiv.org/abs/1612.01022v1 Z. Zhao, W. Chen, X. Wu, P. C. Y. Chen, and J. Liu, “LSTM network: A deep learning approach for Short-term traffic forecast,” IET Intell. Transp. Syst. , vol. 11, no. 2, pp. 68–75, Mar. 2017, doi: 10.1049/iet-its.2016.0208. Z. Duan et al. , “Prediction of city-scale dynamic taxi origin-destination flows using a hybrid deep neural network combined with travel time,” IEEE Access , vol. 7, pp. 127816–127832, 2019, doi: 10.1109/ACCESS.2019.2939902. G. S. Kashyap, A. E. I. Brownlee, O. C. Phukan, K. Malik, and S. Wazir, “Roulette-Wheel Selection-Based PSO Algorithm for Solving the Vehicle Routing Problem with Time Windows,” Jun. 2023, Accessed: Jul. 04, 2023. [Online]. Available: https://arxiv.org/abs/2306.02308v1 P. Kaur, G. S. Kashyap, A. Kumar, M. T. Nafis, S. Kumar, and V. Shokeen, “From Text to Transformation: A Comprehensive Review of Large Language Models’ Versatility,” Feb. 2024, Accessed: Mar. 21, 2024. [Online]. Available: https://arxiv.org/abs/2402.16142v1 M. Kanojia, P. Kamani, G. S. Kashyap, S. Naz, S. Wazir, and A. Chauhan, “Alternative Agriculture Land-Use Transformation Pathways by Partial-Equilibrium Agricultural Sector Model: A Mathematical Approach,” Aug. 2023, Accessed: Sep. 16, 2023. [Online]. Available: https://arxiv.org/abs/2308.11632v1 S. Wazir, G. S. Kashyap, and P. Saxena, “MLOps: A Review,” Aug. 2023, Accessed: Sep. 16, 2023. [Online]. Available: https://arxiv.org/abs/2308.10908v1 G. S. Kashyap et al. , “Revolutionizing Agriculture: A Comprehensive Review of Artificial Intelligence Techniques in Farming,” Feb. 2024, doi: 10.21203/RS.3.RS-3984385/V1. S. Naz and G. S. Kashyap, “Enhancing the predictive capability of a mathematical model for pseudomonas aeruginosa through artificial neural networks,” Int. J. Inf. Technol. 2024 , pp. 1–10, Feb. 2024, doi: 10.1007/S41870-023-01721-W. N. Marwah, V. K. Singh, G. S. Kashyap, and S. Wazir, “An analysis of the robustness of UAV agriculture field coverage using multi-agent reinforcement learning,” Int. J. Inf. Technol. , vol. 15, no. 4, pp. 2317–2327, May 2023, doi: 10.1007/s41870-023-01264-0. A. Nigam and S. Srivastava, “Hybrid deep learning models for traffic stream variables prediction during rainfall,” Multimodal Transp. , vol. 2, no. 1, p. 100052, Mar. 2023, doi: 10.1016/j.multra.2022.100052. Additional Declarations The authors declare no competing interests. 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-5389235","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":374078918,"identity":"4c6ffba9-6e78-401f-ac67-61b7403ac5b6","order_by":0,"name":"GOPICHAND BANDARUPALLI","email":"data:image/png;base64,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","orcid":"","institution":"campbellsville university","correspondingAuthor":true,"prefix":"","firstName":"GOPICHAND","middleName":"","lastName":"BANDARUPALLI","suffix":""}],"badges":[],"createdAt":"2024-11-04 15:12:58","currentVersionCode":1,"declarations":{"humanSubjects":false,"vertebrateSubjects":false,"conflictsOfInterestStatement":false,"humanSubjectEthicalGuidelines":false,"humanSubjectConsent":false,"humanSubjectClinicalTrial":false,"humanSubjectCaseReport":false,"vertebrateSubjectEthicalGuidelines":false},"doi":"10.21203/rs.3.rs-5389235/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-5389235/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":68336964,"identity":"2e1bf112-547e-4d60-aaad-19794755cdb4","added_by":"auto","created_at":"2024-11-06 08:16:39","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":8868,"visible":true,"origin":"","legend":"\u003cp\u003eError distribution for each model\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-5389235/v1/527cd10027b229508cb6e27d.png"},{"id":68336962,"identity":"6f63b621-7d48-44ae-ac1d-7c82cbba09e9","added_by":"auto","created_at":"2024-11-06 08:16:39","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":55114,"visible":true,"origin":"","legend":"\u003cp\u003eRF actual vs predicted values\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-5389235/v1/d1de1953cab32d8b218347e4.png"},{"id":68336965,"identity":"6fc57bb5-49dc-42fd-869f-aaf80b0e0ec4","added_by":"auto","created_at":"2024-11-06 08:16:39","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":56194,"visible":true,"origin":"","legend":"\u003cp\u003eGRU actual vs predicted values\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-5389235/v1/90086929465972f410c55dce.png"},{"id":68338203,"identity":"68324e27-8197-454c-beef-4ba1f16f85da","added_by":"auto","created_at":"2024-11-06 08:24:39","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":56314,"visible":true,"origin":"","legend":"\u003cp\u003eLSTM actual vs predicted values\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-5389235/v1/78481f12cbe8792f7e015e5e.png"},{"id":68338842,"identity":"70d5e574-60b4-4f8e-9a30-ec0fc1223be8","added_by":"auto","created_at":"2024-11-06 08:32:43","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":669467,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5389235/v1/12e77926-486b-45dc-b123-7e5fc233d566.pdf"}],"financialInterests":"The authors declare no competing interests.","formattedTitle":"\u003cp\u003eAdvancing Smart Transportation via AI for Sustainable Traffic Solutions in Saudi Arabia\u003c/p\u003e","fulltext":[{"header":"I.\tINTRODUCTION","content":"\u003cp\u003eSaudi Arabia's Vision 2030 focuses on self-driving cars, which shows how important changing transportation is. To support this vision, a smart transportation system needs to be set up quickly. This system can help deal with traffic jams effectively. During Hajj and Umrah seasons, millions of people visit the Holy cities and other major towns. This causes traffic jams in those areas and nearby places. These traffic problems are in line with the Kingdom's Sustainable Development Goals (SDGs). The main aim of these goals is to improve people's lives by managing movement better. This smart transportation system will predict where traffic jams might happen. This will help to regulate traffic better and reduce accidents, injuries, and the time it takes to travel. It will make life better for everyone. Also, when there is a lot of traffic, more energy gets used and more pollution happens. So, having a system that predicts traffic jams is more important than ever. Artificial Intelligence (AI) methods will be used to create a strong transportation model that can predict where congestion might happen accurately\u0026nbsp;[1], [2], [3], [4], [5], [6]. This system should be able to detect unexpected events like roadblocks or accidents. It can then control them well and keep traffic moving smoothly, especially during busy times. More research will be done to make sure the predictions are accurate, and the system works efficiently. This will help reduce traffic jams and their problems. The whole idea supports the country's Sustainable Development Goals (SDGs), which also focus on having strong infrastructure. This sets the stage for a transportation network that is good for the environment and makes everything move faster because all the parts work well together.\u003c/p\u003e\n\u003cp\u003eThe study is structured into five main sections. Firstly, it provides background information to set the context for the research. Secondly, it reviews related works to highlight existing knowledge and gaps in the field. Thirdly, it describes the materials and methods used in the study, detailing the approach taken to gather and analyse data. Fourthly, it presents the results of the experimental analysis, showcasing the findings and any significant observations made during the research process. Finally, it concludes the study by summarizing key findings and offering suggestions for future research directions and areas of improvement in the field.\u003c/p\u003e"},{"header":"II.\tBACKGROUND","content":"\u003cp\u003eAround the world, many cities have a big problem with traffic. This means roads often have too many cars causing them to move slowly or stop a lot especially when lots of people are going to or from work. There are many reasons for this like more people moving to cities, more people being born, also the economy growing and not enough good roads and public transport. This problem isn't just annoying; it can cause protests, make businesses lose money, and harm nature. It doesn't only affect the city where it happens, but also nearby places. It touches every part of life in a society. Traffic jams in streets, busy city centers, or crowded road crossings are all signs of traffic congestion. It happens when there are too many people trying to use the roads and there are not enough ways for them to travel easily. This happens because of things like where people live and work, how many cars there are, how easy it is to use buses and trains and how the cities are designed. This problem affects healthcare, protecting the environment, and making the economy better. Spending a long time in traffic and using more fuel makes the air dirty and loud. This makes life harder for poor people and makes it tough for them to get jobs, healthcare, and education. Traffic jams also make these businesses lose money because they cannot work well, waste fuel and make it cost more to move things around. To fix this, cities need to plan better for transport in a way that helps everyone and does not hurt the environment.\u003c/p\u003e \u003cp\u003eMany cities worldwide are trying different ways to solve this problem of too many cars on the roads. Several cities are also spending a lot of money on mass transit systems. Others are building more sidewalks for people to walk on. This is to make it harder for people to use their own cars which can cause traffic jams, especially during busy times like rush hours in downtown areas. Another idea is for governments to charge money for driving on certain roads at certain times. This is called congestion charging. It is signified to make people think twice about driving during busy times. Some cities are also using smart technology to control traffic lights which change depending on how many cars are on the road. They can help cars avoid getting stuck in traffic jams. Technology is also helping these cities to manage traffic better. With big data and live monitoring, cities can see what is happening on the roads in real-time. This helps them make decisions to keep traffic moving. It also helps them respond quickly if something goes wrong, like a car breaking down. All these efforts aim to make traffic flow better and reduce the chances of accidents. By encouraging people to use public transportation, walk, or carpool, cities hope to ease congestion and make streets safer for everyone. They also want to make sure that important areas, like hospitals and schools, stay accessible even when there's a lot of traffic.\u003c/p\u003e"},{"header":"III. RELATED WORKS","content":"\u003cp\u003eModel-driven prediction approaches have been extensively utilized in the examination and projection of traffic flow dynamics. These methodologies incorporate various techniques such as moving averages [7], exponential smoothing [8], the Kalman filter [9], and Autoregressive Moving Averages (ARIMA) models [10]. For instance, [11] implemented the exponential smoothing technique to analyze and forecast traffic flow within urban transportation networks. [12] built upon this groundwork by improving a spatio-temporal ARIMA model, customizing it for the accurate forecasting of traffic flow in urban environments at five-minute intervals. Similarly, [13] merged the ARIMA model with Support Vector Regression (SVR), thus enhancing the precision of traffic prediction efforts. Additionally, [14] introduced a non-parametric K-Nearest Neighbor (KNN) model specifically tailored for predicting urban road traffic speed and flow patterns. Finally, [15] developed a dynamic multi-interval traffic flow prediction approach based on KNN non-parametric regression models. These research endeavors collectively highlight the effectiveness and adaptability of model-driven prediction strategies in improving our comprehension and forecasting capabilities within the domain of traffic flow analysis.\u003c/p\u003e \u003cp\u003eHowever, in comparison to model-driven methodologies, data-driven strategies employing Artificial Neural Network (ANN) have demonstrated their effectiveness in predicting traffic flow such as [16], [17]. [18] proposed a traffic flow prediction model based on Multi-Layer Perceptron (MLP) neural networks. [19] employed an ANN to anticipate short-term traffic flow, utilizing factors like traffic volume, speed, density, and time. [20] utilized DL methods to capture complex traffic flow patterns without prior explicit knowledge. Following this, [21] introduced a technique utilizing Convolutional Neural Networks (CNNs) for estimating traffic flow speed across extensive networks. They verified the efficacy of CNNs in extracting localized features through convolutional filter layers, thus mimicking spatial dependencies within urban areas using sliding windows. [17] devised a comprehensive DL framework merging residual and convolutional networks. This framework integrated traffic flow data and weather conditions as inputs, facilitating the capture of temporal similarities across various road segments and enhancing traffic flow prediction accuracy. Additionally, [22] introduced a hybrid DL approach for short-term traffic flow prediction. Furthermore, [23] introduced Long Short-Term Memory (LSTM) neural networks for estimating traffic flow within a traffic network's domain.\u003c/p\u003e \u003cp\u003eRecent studies have also explored various methods to analyze and predict commuter travel patterns. For instance, [24] utilized LSTM networks to estimate commuter travel time, while [10] integrated linear regression with LSTM for forecasting travel demand in urban settings like New York. However, these methods often overlook the combined spatial and temporal aspects of city-scale traffic flow, which are crucial for precise predictions. CNN are adept at capturing local spatial features, whereas LSTM networks excel in capturing long-term dependencies in sequential data [25], [26], [27], [28], [29], [30], [31]. The COVID-19 pandemic has significantly influenced travel behavior, prompting innovative applications of CNN and LSTM networks. Moreover, [32] employed this hybrid model to estimate traffic flow during rainfall events. Building upon these insights, our research introduces a tailored data-driven model to analyze the spatial and temporal aspects of city-scale traffic flow, with a specific focus on short-term predictions. A summary of the studies listed above is presented in Table I.\u003c/p\u003e \u003cp\u003eTABLE I\u003c/p\u003e \u003cp\u003eCOMMON FEATURES IN TRAFFIC FLOW PREDICTION STUDIES\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"No\" id=\"Taba\" border=\"1\"\u003e \u003ccolgroup cols=\"10\"\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 \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c9\" colnum=\"9\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c10\" colnum=\"10\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eStudy\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTime Series Data\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eSpatial Data\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eTemporal Data\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eShort-Term Prediction\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eLong-Term Prediction\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003eHybrid Model\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c8\"\u003e \u003cp\u003eNon-Parametric Model\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c9\"\u003e \u003cp\u003eWeather Consideration\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c10\"\u003e \u003cp\u003eCOVID-19 Impact\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e[7]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e[8]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e[9]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e[10]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e[11]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e[12]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e[13]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e[14]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e[15]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e[16]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e[17]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e[18]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e[19]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e[20]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e[21]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e[22]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e[23]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e[24]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e[32]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e❌\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e✔\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\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"},{"header":"IV. MATERIALS AND METHODS","content":"\u003cp\u003eA. \u003cem\u003eDataset Analysis\u003c/em\u003e\u003c/p\u003e \u003cp\u003eUrban areas are complex systems when it comes to traffic, and they change dramatically throughout different days and times according to data analysis. What we call weekdays have extremely high volumes of traffic in specific time brackets usually called rush hours. This period often starts at 6 am to 8 am and from 4 pm to 6 pm in the evening whereby most people are either going or coming from work thereby heavily impacting on congestion levels. Fridays show unique deviation from what would otherwise be anticipated as typical weekday traffic patterns; despite being considered part of weekdays, congestion is not experienced during normal working hours but there is still increased road usage though not as much as other days of the week due to leisure activities and social gatherings. Monday is the only day that commuters commute consistently throughout these days because it is their first day back at work after taking a weekend off. Therefore, temporal factors should not be disregarded when designing effective management systems since they offer important insights into the potential intensity of bottlenecks at specific times if nothing is done to alleviate them. The dataset description used in this study is compiled in Table II.\u003c/p\u003e \u003cp\u003eTABLE II\u003c/p\u003e \u003cp\u003eSUMMARY OF TEMPORAL TRAFFIC DATASET\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"No\" id=\"Tabb\" border=\"1\"\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\"\u003e \u003cp\u003eDay\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTime Period\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eTraffic Volume\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eKey Observations\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eWeekdays\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e06:00\u0026ndash;08:00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eHigh\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eMorning rush hour due to work commutes\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e16:00\u0026ndash;18:00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eHigh\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eEvening rush hour due to work commutes\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eFriday\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e06:00\u0026ndash;08:00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eLower\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eReduced morning congestion compared to other weekdays\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e16:00\u0026ndash;18:00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eModerate\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eEvening traffic due to social and recreational activities\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWeekends\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eVarious\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eVariable\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eMuch lesser uniformed traffic patterns contrast to weekdays\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\u003eB. \u003cem\u003eModel Analysis\u003c/em\u003e\u003c/p\u003e \u003cp\u003eTemporal traffic pattern insights are incredibly important to urban planners and decision-makers in their effort to reduce congestion and plan for infrastructural growth as well as transport system development towards building resilient cities that can withstand shocks such as earthquakes, among others. There were three algorithms used for predicting future transportation state which are Random Forests (RF), Gated Recurrent Units (GRUs) and Long Short-Term Memories (LSTMs). Among the performance metrics used to evaluate the RF model with 100 estimators included Mean Squared Error (MSE), Root Mean Square Error (RMSE), Mean Absolute Error(MAE), Mean Absolute Percentage Error(MAPE) ,Error cost, Outlier sensitivity, Model complexity etc then compared forecast against actual data for visualizing how good was our model? Like this, the GRU model, which consisted of a single GRU layer followed by two dense layers, underwent extensive training and evaluation. Performance measures were calculated, and predictions were then compared to the real data. However, a type error occurred during the visualization step because the test set, and predictions had different formats. This issue was resolved through transforming the forecasts into a numpy array. The training, assessment, and presentation procedures for the LSTM model\u0026mdash;which consists of one LSTM layer followed by two dense layers\u0026mdash;should be mentioned, among other things. It is also important to note that Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e displays the graphical representation errors distribution for each model, providing some insight into the distribution and concentration sections where such errors in prediction cluster.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e"},{"header":"V.\tEXPERIMENTAL ANALYSIS","content":"\u003cp\u003eIn this section, we discuss traffic forecast techniques using three models: RF, GRU, and LSTM. To enable the chosen models to use the dataset, we first do considerable pre-processing on it. These procedures include encoding categorical information like the day of the week and traffic conditions, as well as standardizing time to a 24-hour format. The data set is split into training and testing sets for model training and evaluation, accordingly, following the pre. Our investigation begins with the RF model, which is well known for its ability to handle complex information. The RF model offers insights into its predicted performance through an extensive evaluation process. Specifically, its ensemble learning approach exhibits competitive performance metrics, indicating strong performance in tasks such as traffic prediction where multiple factors interact to influence the result. The metrics that were employed by the RF model are summarized in Table III and Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e. The mean square error (MSE) for the RF model between the actual and projected values is 15.6. RMSE serves as a valuable metric for assessing the magnitude of error in predictions. It is an approach to understand how far off predictions are from real values. RMSE is calculated by taking the square root of MSE, which stands for Mean Square Error. When RMSE is low, the predictions are generally close to the actual values. So, the relatively low MSE indicates a level of reliability in the model's predictions. For example, if a model has a mean absolute error of 2.75, it means that, on average, the predictions are off by about 2.75 units. This shows that the model makes modest mistakes in its predictions, which is a good sign for its reliability. Another measure that can be looked at is MAPE or Mean Absolute Percentage Error. A MAPE of 5.3% means that, on average, the predictions are within 5.3% of the actual values. This is an important strategy in traffic forecasting because it tells us that the model's predictions are generally pretty close to reality. In real-world traffic forecasting, it is often hard to get detailed or reliable historical data to base predictions on. So, having a model that can make accurate predictions even with such limited data is a crucial role. While managing traffic in cities, it is okay to make some mistakes in predictions because they usually don't cause a big problem. Mistakes might mean a slight delay in traffic plans, but they're not usually too serious. The costs of these mistakes are manageable and can be dealt with by taking suitable actions at strategic points along important paths in the city. One of the reasons why the RF model, or Random Forest model, is good for traffic forecasting is because they are not easily affected by the outliers. These Outliers are unusual or abnormal data points that can sometimes throw off predictions. But the RF model is designed to handle these kinds of irregularities. It is like having a traffic manager who can deal with unexpected events like accidents or sudden increases in traffic volume even without getting confused. This means that the model can make accurate predictions even while facing random data points. The RF model works by building many decision trees and then combining them together. This makes them really good at understanding complex relationships between different factors that affect traffic. It is like having a team of experts who know all the ins and outs of the traffic patterns. However, because the model is so complex, it needs a lot of processing power to work effectively. Overall, the RF model is effective in traffic prediction because it's accurate and reliable. It can handle moderate errors without causing significant problems, making it suitable for real-world scenarios where some errors are expected. It is not easily affected by outliers; it can make accurate predictions even when faced with abnormal data points. So, if there's a need for a model that can make accurate predictions for traffic forecasting, the RF model is a good option to consider. This is especially true if the data points are frequently found along major highways with numerous entrances and exits close to one another over short distances, heavy traffic during peak hours, and sharp changes over time due to various factors like accidents, road works, etc.\u003c/p\u003e \u003cp\u003eTABLE III\u003c/p\u003e \u003cp\u003eRANDOM FOREST MODEL PERFORMANCE METRICS\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"No\" id=\"Tabc\" border=\"1\"\u003e \u003ccolgroup cols=\"2\"\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 \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMetric\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eValue\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMean Squared Error (MSE)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e15.6\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRoot Mean Squared Error (RMSE)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e3.95\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMean Absolute Error (MAE)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2.75\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMean Absolute Percentage Error (MAPE)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e5.3%\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eError Cost\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eModerate\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eOutlier Sensitivity\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eLow\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eModel Complexity\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eHigh\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\u003e \u003c/p\u003e \u003cp\u003eNext, our focus shifts to Recurrent Neural Network (RNN) architectures, beginning with the GRU model. By leveraging its ability to capture sequential dependencies in data, the GRU model is trained and assessed. Despite showing promising results, with significant improvements over traditional ML methods, it does not achieve optimal performance metrics compared to the RF model. Table IV outlines the performance characteristics of the GRU model and provides valuable comparisons with the RF model previously evaluated as depicted in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e. The mean squared variance between the expected and actual values compared to RF is less than the MSE of 12.8, which indicates that the GRU model can capture temporal correlations in traffic data. The model's RMSE of 3.58, which places it higher in terms of overall predictive performance than the RF model, further demonstrates its capacity to foresee with a smaller margin of inaccuracy. The mean absolute error (MAE) of the GRU model is 2.45, which is a lower value and highlights the model's accuracy in predicting traffic patterns. Furthermore, with a MAPE of 4.7% indicating that it is within 4.7% of the real values, the GRU model performs somewhat better than the RF model. Despite these promising metrics, the GRU model has moderate error costs, similar to the RF model, indicating that even though it makes generally accurate predictions, prediction errors can still have adverse operational and economic consequences that must be managed in real-world applications. While handling irregular data points better than many traditional ML models, the GRU model is not as robust as the RF model due to its moderate sensitivity to outliers. The model's performance in scenarios where anomalies occur frequently, such as traffic accidents or sudden increases in volume, may be affected by this moderate sensitivity. In terms of model complexity, the GRU is classified as medium. Due to its gating mechanisms and sequential nature, it is more complex by nature than typical ML models, but not as complex as the ensemble-based RF model. This medium complexity is a suitable option for capturing temporal trends without unduly straining computational resources as it balances computational needs and predictive capabilities.\u003c/p\u003e \u003cp\u003eTABLE IV\u003c/p\u003e \u003cp\u003eGRU MODEL PERFORMANCE METRICS\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"No\" id=\"Tabd\" border=\"1\"\u003e \u003ccolgroup cols=\"2\"\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 \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMetric\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eValue\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMean Squared Error (MSE)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e12.8\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRoot Mean Squared Error (RMSE)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e3.58\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMean Absolute Error (MAE)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2.45\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMean Absolute Percentage Error (MAPE)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e4.7%\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eError Cost\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eModerate\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eOutlier Sensitivity\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eModerate\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eModel Complexity\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMedium\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\u003e \u003c/p\u003e \u003cp\u003eNext, we introduced the LSTM model that has been known for a long time as the best in long-term dependency recognition in sequence data. In this light, traffic prediction tasks are conducted to see how well it can perform. However, there should be some more tests against an RF model, so we know what works better. Known for its depth and memory cells with specialized functions, the LSTM model has proved effective in traffic prediction. The training phase had several other models whose performance metrics were not as good as those of this one since during evaluation it achieved the lowest test loss ever recorded among all considered models. This demonstrates that the only algorithm capable of making such complex predictions about traffic patterns would have been the Long Short-Term Memory (LSTM) model, which processes inputs over time steps into outputs across variable sequence lengths until convergence on some fixed-point. What more could you ask for from a Long Short-Term Memory (LSTM) model? Furthermore, the results obtained from examining the plots depicted in Table V and Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e provide us with indications regarding the potential success or failure of a certain predictive capability in comparison to other similar ones, such as the two displayed here, where they differ. The difference between the two models' accuracy in predicting all points examined so far throughout our research into each model's strengths and weaknesses is ΔY (Actual \u0026ndash; Predicted), which is always within ΔX rather than zero. This indicates that both models perform equally poorly in predicting weak areas closer to either end point, possibly in part as none of them recognize features outside a certain range of values. As an illustration of the lack of consistency between anticipated forecasts made based solely on this type of proof, we can see that, between various points along the x-axis, the most faraway ones are more closely related than the two most adjacent indicated values themselves farther apart, but never precisely identical distance away from each other. This still fails to sufficiently account for the least squares fits noticed.\u003c/p\u003e \u003cp\u003eTABLE V\u003c/p\u003e \u003cp\u003eLSTM MODEL PERFORMANCE METRICS\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"No\" id=\"Tabe\" border=\"1\"\u003e \u003ccolgroup cols=\"2\"\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 \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMetric\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eValue\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMean Squared Error (MSE)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e10.5\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRoot Mean Squared Error (RMSE)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e3.24\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMean Absolute Error (MAE)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2.15\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMean Absolute Percentage Error (MAPE)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e3.9%\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eError Cost\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eLow\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eOutlier Sensitivity\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eLow\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eModel Complexity\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eHigh\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\u003e \u003c/p\u003e \u003cp\u003eThe experimental investigation showed that a variety of traffic forecast models, ranging from more sophisticated deep learning (DL) structures like Gated Recurrent Units (GRU) and Long Short-Term Memory (LSTM) to more conventional techniques like Random Forest (RF), are effective. Each model in the realm of traffic prediction has demonstrated pros and cons of its own. However, the LSTM model outperformed the others, achieving the highest accurate rate in traffic trend predicting. These results are critical to transportation planning and management because they offer practical guidance on enhancing system efficiency and traffic flow optimization.\u003c/p\u003e"},{"header":"VI. CONCLUSION AND FUTURE WORKS ","content":"\u003cp\u003eIn general, our test indicates the many traffic prediction methods that can be used. The study shows that both regular machine learning and deep learning are good at predicting things. We looked at RF, GRU, and LSTM models and found them to be strong. Among them, the LSTM model looks especially promising. It's good at understanding long-term connections between traffic data and complicated patterns. These findings matter a lot for planning and managing transportation. They help make traffic flow better and systems more efficient. In the future, we should focus on making DL models like LSTM even better. We can also mix them with other methods to make them stronger. This indicates that using real-time data to make predictions is more accurate and we need to make sure these predictions work well in real life, especially in big cities with lots of traffic. Additionally, they will investigate how things like weather and the city changes affect traffic. This will help us to make even better models that can handle the challenges of city life.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eA. Funding:\u0026nbsp;\u003c/strong\u003eNo funds, grants, or other support was received.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eB. Conflict of Interest:\u0026nbsp;\u003c/strong\u003eThe authors declare that they have no known competing for financial interests or personal relationships that could have appeared to influence the work reported in this paper.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eC. Data Availability:\u0026nbsp;\u003c/strong\u003eData will be made on reasonable request.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eD. Code Availability:\u0026nbsp;\u003c/em\u003e\u003c/strong\u003eCode will be made on reasonable request.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n \u003cli\u003eH. Habib, G. S. Kashyap, N. Tabassum, and T. Nafis, \u0026ldquo;Stock Price Prediction Using Artificial Intelligence Based on LSTM\u0026ndash; Deep Learning Model,\u0026rdquo; in \u003cem\u003eArtificial Intelligence \u0026amp; Blockchain in Cyber Physical Systems: Technologies \u0026amp; Applications\u003c/em\u003e, CRC Press, 2023, pp. 93\u0026ndash;99. doi: 10.1201/9781003190301-6.\u003c/li\u003e\n \u003cli\u003eS. Wazir, G. S. Kashyap, K. Malik, and A. E. I. Brownlee, \u0026ldquo;Predicting the Infection Level of COVID-19 Virus Using Normal Distribution-Based Approximation Model and PSO,\u0026rdquo; Springer, Cham, 2023, pp. 75\u0026ndash;91. doi: 10.1007/978-3-031-33183-1_5.\u003c/li\u003e\n \u003cli\u003eG. S. Kashyap, D. Mahajan, O. C. Phukan, A. Kumar, A. E. I. Brownlee, and J. Gao, \u0026ldquo;From Simulations to Reality: Enhancing Multi-Robot Exploration for Urban Search and Rescue,\u0026rdquo; Nov. 2023, Accessed: Dec. 03, 2023. [Online]. Available: https://arxiv.org/abs/2311.16958v1\u003c/li\u003e\n \u003cli\u003eG. S. Kashyap, A. Siddiqui, R. Siddiqui, K. Malik, S. Wazir, and A. E. I. Brownlee, \u0026ldquo;Prediction of Suicidal Risk Using Machine Learning Models,\u0026rdquo; Dec. 25, 2021. Accessed: Feb. 04, 2024. [Online]. Available: https://papers.ssrn.com/abstract=4709789\u003c/li\u003e\n \u003cli\u003eG. S. Kashyap, K. Malik, S. Wazir, and R. Khan, \u0026ldquo;Using Machine Learning to Quantify the Multimedia Risk Due to Fuzzing,\u0026rdquo; \u003cem\u003eMultimed. Tools Appl.\u003c/em\u003e, vol. 81, no. 25, pp. 36685\u0026ndash;36698, Oct. 2022, doi: 10.1007/s11042-021-11558-9.\u003c/li\u003e\n \u003cli\u003eG. S. Kashyap \u003cem\u003eet al.\u003c/em\u003e, \u0026ldquo;Detection of a facemask in real-time using deep learning methods: Prevention of Covid 19,\u0026rdquo; Jan. 2024, Accessed: Feb. 04, 2024. [Online]. Available: https://arxiv.org/abs/2401.15675v1\u003c/li\u003e\n \u003cli\u003eQ. Liu, E. Chung, and L. Zhai, \u0026ldquo;Fusing moving average model and stationary wavelet decomposition for automatic incident detection: case study of Tokyo Expressway,\u0026rdquo; \u003cem\u003eJ. Traffic Transp. Eng. (English Ed.\u003c/em\u003e, vol. 1, no. 6, pp. 404\u0026ndash;414, Dec. 2014, doi: 10.1016/S2095-7564(15)30290-7.\u003c/li\u003e\n \u003cli\u003eK. Y. Chan, T. S. Dillon, J. Singh, and E. Chang, \u0026ldquo;Traffic flow forecasting neural networks based on exponential smoothing method,\u0026rdquo; in \u003cem\u003eProceedings of the 2011 6th IEEE Conference on Industrial Electronics and Applications, ICIEA 2011\u003c/em\u003e, 2011, pp. 376\u0026ndash;381. doi: 10.1109/ICIEA.2011.5975612.\u003c/li\u003e\n \u003cli\u003eS. V. Kumar, \u0026ldquo;Traffic Flow Prediction using Kalman Filtering Technique,\u0026rdquo; in \u003cem\u003eProcedia Engineering\u003c/em\u003e, No longer published by Elsevier, Jan. 2017, pp. 582\u0026ndash;587. doi: 10.1016/j.proeng.2017.04.417.\u003c/li\u003e\n \u003cli\u003eT. Mai, B. Ghosh, and S. Wilson, \u0026ldquo;Short-term traffic-flow forecasting with auto-regressive moving average models,\u0026rdquo; \u003cem\u003eProc. Inst. Civ. Eng. Transp.\u003c/em\u003e, vol. 167, no. 4, pp. 232\u0026ndash;239, May 2014, doi: 10.1680/tran.12.00012.\u003c/li\u003e\n \u003cli\u003eB. M. Williams, P. K. Durvasula, and D. E. Brown, \u0026ldquo;Urban freeway traffic flow prediction: Application of seasonal autoregressive integrated moving average and exponential smoothing models,\u0026rdquo; \u003cem\u003eTransp. Res. Rec.\u003c/em\u003e, no. 1644, pp. 132\u0026ndash;141, Jan. 1998, doi: 10.3141/1644-14.\u003c/li\u003e\n \u003cli\u003eQ. Ding, X. Wang, X. Zhang, and Z. Sun, \u0026ldquo;Forecasting traffic volume with space-time ARIMA model,\u0026rdquo; in \u003cem\u003eAdvanced Materials Research\u003c/em\u003e, Trans Tech Publications Ltd, 2011, pp. 979\u0026ndash;983. doi: 10.4028/www.scientific.net/AMR.156-157.979.\u003c/li\u003e\n \u003cli\u003eL. Li, S. He, J. Zhang, and B. Ran, \u0026ldquo;Short-term highway traffic flow prediction based on a hybrid strategy considering temporal\u0026ndash;spatial information,\u0026rdquo; \u003cem\u003eJ. Adv. Transp.\u003c/em\u003e, vol. 50, no. 8, pp. 2029\u0026ndash;2040, Dec. 2016, doi: 10.1002/atr.1443.\u003c/li\u003e\n \u003cli\u003eD. Xia, B. Wang, H. Li, Y. Li, and Z. Zhang, \u0026ldquo;A distributed spatial-temporal weighted model on MapReduce for short-term traffic flow forecasting,\u0026rdquo; \u003cem\u003eNeurocomputing\u003c/em\u003e, vol. 179, pp. 246\u0026ndash;263, Feb. 2016, doi: 10.1016/j.neucom.2015.12.013.\u003c/li\u003e\n \u003cli\u003eH. Chang, Y. Lee, B. Yoon, and S. Baek, \u0026ldquo;Dynamic near-term traffic flow prediction: System-oriented approach based on past experiences,\u0026rdquo; \u003cem\u003eIET Intell. Transp. Syst.\u003c/em\u003e, vol. 6, no. 3, pp. 292\u0026ndash;305, Sep. 2012, doi: 10.1049/iet-its.2011.0123.\u003c/li\u003e\n \u003cli\u003eT. Kim, S. Sharda, X. Zhou, and R. M. Pendyala, \u0026ldquo;A stepwise interpretable machine learning framework using linear regression (LR) and long short-term memory (LSTM): City-wide demand-side prediction of yellow taxi and for-hire vehicle (FHV) service,\u0026rdquo; \u003cem\u003eTransp. Res. Part C Emerg. Technol.\u003c/em\u003e, vol. 120, p. 102786, Nov. 2020, doi: 10.1016/j.trc.2020.102786.\u003c/li\u003e\n \u003cli\u003eJ. Zhang, Y. Zheng, and D. Qi, \u0026ldquo;Deep spatio-temporal residual networks for citywide crowd flows prediction,\u0026rdquo; in \u003cem\u003e31st AAAI Conference on Artificial Intelligence, AAAI 2017\u003c/em\u003e, AAAI press, Feb. 2017, pp. 1655\u0026ndash;1661. doi: 10.1609/aaai.v31i1.10735.\u003c/li\u003e\n \u003cli\u003eZ. Zhu, B. Peng, C. Xiong, and L. Zhang, \u0026ldquo;Short-term traffic flow prediction with linear conditional Gaussian Bayesian network,\u0026rdquo; \u003cem\u003eJ. Adv. Transp.\u003c/em\u003e, vol. 50, no. 6, pp. 1111\u0026ndash;1123, Oct. 2016, doi: 10.1002/atr.1392.\u003c/li\u003e\n \u003cli\u003eK. Kumar, M. Parida, and V. K. Katiyar, \u0026ldquo;Short Term Traffic Flow Prediction for a Non Urban Highway Using Artificial Neural Network,\u0026rdquo; \u003cem\u003eProcedia - Soc. Behav. Sci.\u003c/em\u003e, vol. 104, pp. 755\u0026ndash;764, Dec. 2013, doi: 10.1016/j.sbspro.2013.11.170.\u003c/li\u003e\n \u003cli\u003eY. Lv, Y. Duan, W. Kang, Z. Li, and F. Y. Wang, \u0026ldquo;Traffic Flow Prediction with Big Data: A Deep Learning Approach,\u0026rdquo; \u003cem\u003eIEEE Trans. Intell. Transp. Syst.\u003c/em\u003e, vol. 16, no. 2, pp. 865\u0026ndash;873, Apr. 2015, doi: 10.1109/TITS.2014.2345663.\u003c/li\u003e\n \u003cli\u003eX. Ma, Z. Tao, Y. Wang, H. Yu, and Y. Wang, \u0026ldquo;Long short-term memory neural network for traffic speed prediction using remote microwave sensor data,\u0026rdquo; \u003cem\u003eTransp. Res. Part C Emerg. Technol.\u003c/em\u003e, vol. 54, pp. 187\u0026ndash;197, May 2015, doi: 10.1016/j.trc.2015.03.014.\u003c/li\u003e\n \u003cli\u003eY. Wu and H. Tan, \u0026ldquo;Short-term traffic flow forecasting with spatial-temporal correlation in a hybrid deep learning framework,\u0026rdquo; Dec. 2016, Accessed: May 22, 2024. [Online]. Available: https://arxiv.org/abs/1612.01022v1\u003c/li\u003e\n \u003cli\u003eZ. Zhao, W. Chen, X. Wu, P. C. Y. Chen, and J. Liu, \u0026ldquo;LSTM network: A deep learning approach for Short-term traffic forecast,\u0026rdquo; \u003cem\u003eIET Intell. Transp. Syst.\u003c/em\u003e, vol. 11, no. 2, pp. 68\u0026ndash;75, Mar. 2017, doi: 10.1049/iet-its.2016.0208.\u003c/li\u003e\n \u003cli\u003eZ. Duan \u003cem\u003eet al.\u003c/em\u003e, \u0026ldquo;Prediction of city-scale dynamic taxi origin-destination flows using a hybrid deep neural network combined with travel time,\u0026rdquo; \u003cem\u003eIEEE Access\u003c/em\u003e, vol. 7, pp. 127816\u0026ndash;127832, 2019, doi: 10.1109/ACCESS.2019.2939902.\u003c/li\u003e\n \u003cli\u003eG. S. Kashyap, A. E. I. Brownlee, O. C. Phukan, K. Malik, and S. Wazir, \u0026ldquo;Roulette-Wheel Selection-Based PSO Algorithm for Solving the Vehicle Routing Problem with Time Windows,\u0026rdquo; Jun. 2023, Accessed: Jul. 04, 2023. [Online]. Available: https://arxiv.org/abs/2306.02308v1\u003c/li\u003e\n \u003cli\u003eP. Kaur, G. S. Kashyap, A. Kumar, M. T. Nafis, S. Kumar, and V. Shokeen, \u0026ldquo;From Text to Transformation: A Comprehensive Review of Large Language Models\u0026rsquo; Versatility,\u0026rdquo; Feb. 2024, Accessed: Mar. 21, 2024. [Online]. Available: https://arxiv.org/abs/2402.16142v1\u003c/li\u003e\n \u003cli\u003eM. Kanojia, P. Kamani, G. S. Kashyap, S. Naz, S. Wazir, and A. Chauhan, \u0026ldquo;Alternative Agriculture Land-Use Transformation Pathways by Partial-Equilibrium Agricultural Sector Model: A Mathematical Approach,\u0026rdquo; Aug. 2023, Accessed: Sep. 16, 2023. [Online]. Available: https://arxiv.org/abs/2308.11632v1\u003c/li\u003e\n \u003cli\u003eS. Wazir, G. S. Kashyap, and P. Saxena, \u0026ldquo;MLOps: A Review,\u0026rdquo; Aug. 2023, Accessed: Sep. 16, 2023. [Online]. Available: https://arxiv.org/abs/2308.10908v1\u003c/li\u003e\n \u003cli\u003eG. S. Kashyap \u003cem\u003eet al.\u003c/em\u003e, \u0026ldquo;Revolutionizing Agriculture: A Comprehensive Review of Artificial Intelligence Techniques in Farming,\u0026rdquo; Feb. 2024, doi: 10.21203/RS.3.RS-3984385/V1.\u003c/li\u003e\n \u003cli\u003eS. Naz and G. S. Kashyap, \u0026ldquo;Enhancing the predictive capability of a mathematical model for pseudomonas aeruginosa through artificial neural networks,\u0026rdquo; \u003cem\u003eInt. J. Inf. Technol. 2024\u003c/em\u003e, pp. 1\u0026ndash;10, Feb. 2024, doi: 10.1007/S41870-023-01721-W.\u003c/li\u003e\n \u003cli\u003eN. Marwah, V. K. Singh, G. S. Kashyap, and S. Wazir, \u0026ldquo;An analysis of the robustness of UAV agriculture field coverage using multi-agent reinforcement learning,\u0026rdquo; \u003cem\u003eInt. J. Inf. Technol.\u003c/em\u003e, vol. 15, no. 4, pp. 2317\u0026ndash;2327, May 2023, doi: 10.1007/s41870-023-01264-0.\u003c/li\u003e\n \u003cli\u003eA. Nigam and S. Srivastava, \u0026ldquo;Hybrid deep learning models for traffic stream variables prediction during rainfall,\u0026rdquo; \u003cem\u003eMultimodal Transp.\u003c/em\u003e, vol. 2, no. 1, p. 100052, Mar. 2023, doi: 10.1016/j.multra.2022.100052.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"Campbellsville University","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":"Artificial Intelligence, Deep Learning, Support Vector Machines, Traffic Congestion, United Nations'","lastPublishedDoi":"10.21203/rs.3.rs-5389235/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-5389235/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThe Saudi Arabian government has committed more than $100 billion (USD) to improving the country's transportation infrastructure, in line with Vision 2030 and the Sustainable Development Goals (SDGs) of the United Nations. The National Center for Transportation Safety (NCTS), which focuses on road safety, and the \"Rental Contracts\" initiative are two examples of the infrastructure development projects for which the FY2022 budget allotted 42 billion SAR. On the other hand, as cities become more populated, traffic congestion has worsened, making living more difficult. In response to these issues, the government is putting in place intelligent transportation systems that use Artificial Intelligence (AI) methods to predict traffic patterns and provide drivers with other routes that cut down on travel time. These AI-driven forecasts are anticipated to lessen traffic-related problems like pollution and health hazards, supporting the country's larger objectives for sustainable infrastructure. AI models, such as Random Forest (RF), Gated Recurrent Units (GRU), and Long Short-Term Memory (LSTM), have been shown to be useful in traffic prediction based on empirical results. With a Mean Square Error (MSE) of 10.5, a Root Mean Square Error (RMSE) of 3.24, a Mean Absolute Error (MAE) of 2.15, and a Mean Absolute Percentage Error (MAPE) of 3.9%, the LSTM model outperformed both the RF and GRU models. These findings demonstrate how AI-driven models may help Saudi Arabia create transportation systems that are reliable, effective, and sustainable.\u003c/p\u003e","manuscriptTitle":"Advancing Smart Transportation via AI for Sustainable Traffic Solutions in Saudi Arabia","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-11-06 08:16:35","doi":"10.21203/rs.3.rs-5389235/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":"5da960ac-304e-4795-a1a2-b25e467d33c1","owner":[],"postedDate":"November 6th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":39799610,"name":"Artificial Intelligence and Machine Learning"}],"tags":[],"updatedAt":"2024-11-06T08:16:35+00:00","versionOfRecord":[],"versionCreatedAt":"2024-11-06 08:16:35","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-5389235","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-5389235","identity":"rs-5389235","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}