Prediction of the Compressive Strength of Jute fibre Reinforced Concrete: A Comparative Study of ANFIS, ANN, RF and RT models

The estimation of compressive strength (CS) of jute fibre reinforced concrete (JFRC) is assessed with Adaptive Neuro-Fuzzy Inference System (ANFIS), Artificial Neural Network (ANN), Random Forest (RF), and Random Tree (RT). The present study determines the best-suited model to estimate the CS of JFRC. A total of 93 experimentation observations were extracted from the literature. 70% of random data was used for training and 30% as testing subsets. Models were formulated using different input combinations i.e., aspect ratio, % of fiber, and no. of curing days to predict the CS of JFRC. Correlation Coefficient (CC), Root Mean Square Error (RMSE), and Mean Absolute Error (MAE) were used to evaluate the performance of formulated models. The results showed that the RF model outperforms when compared with ANFIS, ANN, and RT models with CC (0.984, 0.912), RMSE (1.300, 2.641), and MAE (1.016, 2.162) for the training and testing stage.