Comparative Analysis of Machine Learning Approaches for Optimizing Rainfall Prediction for Enhanced Agricultural Sustainability

Abstract

This study focuses on improving and enhancing agricultural output and reducing the negative effects of unpredictability in the weather on food and water security, the purpose of this study is to develop methods for predicting daily rainfall. The purpose of this research is to discover major atmospheric factors that influence rainfall and to forecast its daily occurrence. The implementation of machine learning and deep learning strategies will be utilized to achieve this goal. Data from Kaggle, specifically the "Rainfall in India from 2005-2020" dataset, is utilized to evaluate the effectiveness of Multivariate Linear Regression, Random Forest, and Artificial Neural Network algorithms. The Pearson correlation technique aids in selecting relevant environmental variables for model inputs. The performance of the model is evaluated using evaluation metrics such as root mean squared error and mean absolute error, and the results indicate that the Artificial Neural Network method has a superior predictive capability. This study not only contributes to a better understanding of the complex atmospheric dynamics that are responsible for the patterns of rainfall, but it also has practical implications for improving the management of water resources and enhancing the resilience of agricultural systems in the face of climate change.