Revolutionizing Public Health Surveillance: Computational Solution for Dengue Prediction

Abstract

Dengue fever is a rapidly growing vector-borne viral illness that is posing a threat to an increasing number of areas worldwide. Numerous scientists have focused on various strategies to stop and limit the spread of illness. Additionally, the development of a variety of methods for ascertaining and predictive modeling through quantifiable, numerical analysis of machine learning (ML) is investigated. This research introduces a novel cubic-kernelized support vector machine with a bee optimizer (CSVM-BO) approach for dengue identification. The climate data is initially collected and preprocessed utilizing the decimal scale normalization method. In addition, we provide a feature extraction approach for linear discriminant analysis (LDA), which attempts to extract essential information for early identification and risk assessment. The findings demonstrate that the framework suggested in this research has considerable benefits in public health for dengue prediction and that our proposed CSVM-BO technique performs optimally in terms of latency (5 s), time complexity (0.62 s), and accuracy (98.57%). The problems identified by this thorough study offer a helpful foundation for epidemiology and public health research.