Optimized Feature-Driven Dengue Diagnosis Using Explainable Machine Learning Approaches

Dengue fever is a mosquito-borne viral infection that causes high fever, severe headache, muscle and joint pain, and skin rashes-often called “breakbone fever” due to its intense symptoms. Early and accurate classification of dengue cases is critical for effective treatment and healthcare planning. This research presents a Machine Learning (ML) approach for classifying dengue-affected individuals using various clinical and demographic parameters. Several types of classification models, such as Logistic Regression (LR), Random Forest Classifier (RFC), Support Vector Machine (SVM) and others. K-Nearest Neighbors (KNN), and Multi-Layer Perceptron Classifier (MLPC), are applied, and the RFC achieves the best performance with accuracy, precision, recall, and F1-score of 99.34%. A feature selection technique, Recursive Feature Elimination (RFE), is incorporated to improve the model's efficiency further, enhancing the RFC model's accuracy to 99.67 %. The results demonstrate a significant advancement over traditional models. Additionally, Explainable Artificial Intelligence (XAI) methods, specifically Local Interpretable Model-agnostic Explanations (LIME) and Shapley Additive Explanations (SHAP), are used to interpret model predictions, highlighting key features influencing the classification. This integrated approach boosts classification accuracy and provides meaningful insights to support healthcare professionals in managing dengue cases more effectively.