Logistic Regression Based Next-Day Rain Prediction Model

Abstract

Rain prediction is challenging due to the complex nonlinear combination of atmospheric factors. This paper presents the application of logistic regression modelling to predict rain the next day using weather parameters from the previous days. One year of weather data (temperature, pressure, humidity, sunshine, evaporation, cloud cover, wind direction, and wind speed) from Canberra, Australia has been used to develop the logistic regression-based model. Akaike Information Criterion (AIC) Backward, Baysian Information Criterion (BIC) Stepwise, and Least Absolute Shrinkage and Selection Operator (LASSO) logistic regression models have been developed based on input variable selection and prediction. These models are evaluated using Area Under the ROC Curve(AUC) and Hosmer-Lemeshow test to determine the models' adequacies and accuracies to predict rainfall occurrence the next day. The likelihood of rainfall the next day has been interpreted based on the calculated odds ratios with 95% confidence intervals of the selected independent weather parameters. The result showed that the rainfall the next day can be predicted using logistic regression (AIC Backward) with 87% accuracy, provided that the appropriate weather parameters are chosen.

Machine Learning-Based Rainfall Prediction: Unveiling Insights and Forecasting for Improved Preparedness

“…Ejike et al [9] explored the application of logistic regression modeling to predict rainfall for the following day. They used one year of meteorological data from Canberra, Australia, including temperature, pressure, humidity, sunlight, evaporation, cloud cover, wind direction, and wind speed.…”