International Journal of Environment and Climate Change

Nowadays, many Machine Learning (ML) models gaining prominence due to their accuracy in predicting the time series data. Because of, availability of, different machine learning models, it is important to know the difference between them. A python-based and open-source application, known as Orange software (Version 3.38.1), used to train the ML models. The study was conducted in the Raichur city, Karnataka, India. The data of 45 years were collected which includes a dependent variable i.e., maximum temperature and independent variables i.e., relative humidity, wind speed, surface pressure and precipitation. Six machine learning models namely, Linear Regression (LR), k-Nearest Neighbor (kNN), Support Vector Machine (SVM), Artificial Neural Network (ANN), along with two ensemble models such as Random Forest (RF) and Adaptive Boosting (AB) were selected to analyze the data. Each model was trained by trial and error then the trained model was validated/predicted with the observed values. The seven performance indicators were recorded for each model and compared. Out of six ML models, the ANN model perform better in all seven performance indicators. Specifically, it was found that the R² value of the ANN model in both training and testing stages was  and  respectively. The ANN model occupies rank 1, after comparing six ML models, with  accuracy in validating/predicting the maximum temperature. Hence, the ANN model can be suggested for validation/prediction of maximum temperature around the regions of the Raichur city.