Designing Machine Learning Algorithms for Feature Extraction and Classification in Dengue Diagnosis

In modern healthcare, predicting diseases and identifying their underlying causes are crucial areas of study. This paper proposes a novel feature selection method based on entropy scores and ideal instance ranking to pinpoint key attributes responsible for disease manifestation. By applying this technique to dengue data sourced from various repositories, the study demonstrates the effectiveness of the proposed method in identifying relevant features. The approach utilizes Python IDE for simulation, providing detailed testing and outcome analysis. Results indicate that the feature selection technique outperforms multiple classifiers, achieving higher accuracy and precision. This highlights the potential of the proposed method in improving disease detection and prediction, specifically for dengue, by efficiently selecting informative features from large datasets. The proposed model offers a more accurate, reliable, and effective alternative for identifying disease-related traits compared to conventional classification models, marking a significant step forward in automated disease detection systems.