Integrating Feature Selection with Deep Learning for Enhanced Obesity Prediction Accuracy in Adults

Existing obesity prediction algorithms generally rely on conventional machine learning approaches such as decision trees and support vector machines, which struggle with large datasets and complex, nonlinear relationships, resulting in overfitting and poor accuracy. To overcome these limitations, the proposed approach integrates deep learning models with two complex feature selection techniques, recursive feature elimination (RFE) and L1 regularization. The combination increases the model's capacity to detect complicated patterns while effectively lowering the complexity of the input. Prediction accuracy is greatly increased by the system, and performance measures demonstrate notable improvements over conventional models. To be more precise, the proposed system outperformed the existing systems with an accuracy of 85.7%, precision of 84.5%, recall of 86.3%, and F1-score of 88.8%. Healthcare workers can gain useful insights into predicting and controlling adult obesity with the strong and accurate model.