##plugins.themes.bootstrap3.article.main##

Mohamad Emad Bitar

Dr. V. Sujatha

Abstract

Heat stroke is a severe condition resulting from prolonged exposure to high temperatures, posing significant health risks, particularly during extreme weather events. Accurate prediction of heat stroke is challenging due to the complex interplay of environmental and physiological factors. This study proposes a comprehensive machine learning framework to address this challenge effectively.
We begin by ensuring data quality through normalization using a Reorder Iterative Imputer, which handles missing values and outliers with precision. Feature selection is then performed using a combination of Principal Component Analysis (PCA) and t-Distributed Stochastic Neighbor embedding (t-SNE) to identify the most relevant variables from a diverse set of indicators. The
classification task is executed using a fused machine learning approach, integrating Random Forest, Support Vector Classifier, and Gradient Boosting methods to enhance prediction accuracy. Further, prediction optimization is achieved using a Cascaded Levy Flight Optimization algorithm, finetuning model parameters for superior performance. The proposed method demonstrates significant
improvements in prediction accuracy and reliability over traditional approaches, offering a valuable tool for early intervention and preventive measures in health monitoring and safety management.

##plugins.themes.bootstrap3.article.details##