A Hybrid Model for Rain Prediction Using Machine Learning Algorithm

The objective of this paper is to create an algorithm which performs better than current algorithms for rain prediction on any dataset. The proposed hybrid model is evaluated on the Delhi dataset because it was the only dataset currently easily available and other datasets quality was not promising. This study proposes a hybrid model of combination of Linear Regression (LR) with Random Forest Classifier (RFC) and Gradient Boosting Classifier (GBC) algorithms. The proposed method in this paper has preliminary data processing, feature selection, training RFC and GBC on the training data, combining their outputs, training LR on the combined predictions, and evaluating the final model's performance using metrics such as F1 score, accuracy, precision, and recall. RFC and GBC are decision tree-based algorithms that generates strong and weak decision trees, respectively. The results suggest that the proposed hybrid model outperforms the currently available individual algorithms which has 98% accuracy, providing more accurate rain predictions with 99.9% accuracy. This study’s findings may have practical implications for weather forecasting in Delhi or may use other datasets, which could help government agencies and citizens make informed decisions to mitigate the impacts of rainfall.