A SYSTEMATIC ANALYSIS OF DIFFERENT MACHINE LEARNING & DEEP LEARNING TECHNIQUES USED FOR PREDICTING CYCLONES

Abstract

Cyclones, with their catastrophic potential, pose enormous hazards to coastline communities around the world. Promising outcomes have been observed in the prediction of cyclone behavior through the incorporation of machine-learning techniques in meteorological research in recent years. This research paper presents a systematic analysis of various machine learning methodologies employed for cyclone prediction. This research deals with the collection and pre-processing of extensive meteorological data, spanning historical cyclone occurrences and associated atmospheric conditions. This research paper conducts a comprehensive assessment of different ML methods for cyclone prediction. The main goal is to assess how well these methods work for predicting important cyclone characteristics including intensity, track, and landfall locations. The objective is to shed light on how various algorithms—including SVM, decision trees, neural networks, etc.—perform.