A review on rainfall forecasting using ensemble learning techniques

Abstract

Significant challenges to human health and life have arisen as a result of heavy rains. Floods and other natural disasters that affect people all over the world every year are caused by prolonged periods of heavy rainfall. Predictions of rainfall must be accurate in countries like India where agriculture is the primary occupation. The non-linearity of rainfall makes machine learning (ML) methods more efficient than many other approaches. In machine learning (ML), individual classifiers are less accurate than ensemble learning (EL) techniques. In order to better understand the various Machine Learning algorithms and Ensemble Learning techniques that researchers employ to predict rainfall, this review paper has been written.This article reviews ensemble learning algorithms for predicting rainfall. In order to increase the accuracy of rainfall forecasts and consequently avoid the negative effects of heavy precipitation, ensemble learning algorithms have gained popularity. This review article examines and makes reference to the development of ensemble approaches, including bagging, boosting, and stacking. The findings of this survey demonstrate that ensemble techniques are much superior to conventional (individual) model learning in terms of rainfall prediction. Additionally, boosting techniques (such enabling, AdaBoost, and extreme gradient boosting) have been applied more frequently and successfully in scenarios involving rainfall forecasting.

Keywords

• Rainfall
• ML
• EL
• MLR
• SVM
• ARIMA