Secure Dengue Epidemic Prediction System: Healthcare Perspective

Abstract

Viral diseases transmitted by mosquitoes are emerging public health problems across the globe. Dengue is considered to be the most significant mosquito-oriented disease. Conspicuously, the present study provides an effective architecture for Dengue Virus Infection surveillance. The proposed system involves a 4-level architecture for the prediction and prevention of dengue infection outspread. The architectural levels including Dengue Information Acquisition level, Dengue Information Classification level, Dengue-Mining and Extraction level, and Dengue-Prediction and Decision Modeling level enable an individual to periodically monitor his/her probabilistic dengue fever measure. The prediction process is carried out so that proactive measures are taken beforehand. For predictive purposes, probabilistic analysis in terms of Level of Dengue Fever (LoDF) was carried out using the Adaptive Neuro-Fuzzy Inference System. Based on the Self-Organized Mapping procedure, the presence of LoDF is visualized. Several simulations on datasets of 16 individuals cumulating to 32,255 instances were conducted to test the effectiveness of the presented model. In comparison to other decision-modeling methods, significantly improved results in form of classification efficacy, a temporal delay, prediction effectiveness, reliability, and stability were reported for the presented model.