Optimizing Climate Data Classification with a Novel Ensemble-based Stacking Classifier

This research explores the novel machine learning model using stacking classifier with ensemble learning for climate data classification. The proposed model systematically integrates various machine learning techniques, encompassing crucial phases including data preprocessing, feature engineering, and model stacking. By leveraging a combination of logistic regression, Support Vector Machine (SVM), XGBoost and Random Forest models, this research paper demonstrates how ensemble approaches can effectively capture complex patterns within climate data. The dataset, sourced from the National Oceanic and Atmospheric Administration’s (NOAA) data repository of hurricane data, includes a wide array of features related to hurricanes and tropical storms, making it ideal for analyzing and predicting storm status. The use of a Stacking Classifier combining multiple models (logistic regression, Random Forest, SVM, XGBoost) appears to be quite effective, achieving 95.74% accuracy, outperforming individual models and simpler ensemble methods. The research establishes the prospective of machine learning techniques to improve climate data analysis and prediction accuracy.