Rainfall Prediction Using Machine Learning: A Comparative Analysis of MLR and Artificial Neural Networks

Abstract

Rainfall plays a vital role in the economy of countries where agriculture is the primary source of income, such as 
India. Accurate rainfall forecasting is crucial not only for effective agricultural planning but also for safeguarding 
communities, particularly those in coastal regions, from risks associated with heavy rain and flooding. Early and 
reliable predictions allow individuals to take preventive measures, reducing potential losses and improving overall 
preparedness. 
This study focuses on developing an efficient rainfall prediction model using Multiple Linear Regression (MLR) 
and Artificial Neural Networks (ANN). The research compares the performance of various machine learning 
algorithms, including MLR, Neural Networks, K-means, and Naïve Bayes, to determine the most suitable 
approach for rainfall forecasting. The input dataset integrates multiple meteorological parameters such as 
humidity, temperature (minimum and maximum), pressure, cloud cover, and wind speed to enhance prediction 
accuracy. 
The proposed model is evaluated using metrics like Mean Absolute Error (MAE), accuracy, and correlation. 
Results indicate that the suggested approach outperforms existing methods in terms of precision and reliability, 
making it a promising solution for rainfall prediction and its associated applications in agriculture and disaster 
management.