Prediction of Chikungunya disease using PSO-based adaptive neuro-fuzzy inference system model

Abstract

ABSTRACT Chikungunya is one of the rapidly spreading viruses which is transmitted by infected mosquitoes and is one of the global health issues. As population is growing at fast pace, so as the data related to patients, health staff ,and doctors. Many machine learning models for the classification of Chikungunya infection were proposed earlier, however, majority of these models suffer from hyper-parameter tuning. In this paper, we propose a novel machine learning model in which hyper-parameters of adaptive neuro-fuzzy inference system (ANFIS) model are optimized using crossover-based Particle Swarm Optimization (PSO). To improve the accuracy and performance of classification, the ANFIS model is optimized using crossover-based PSO. Then, this ANFIS model is trained and tested on the given dataset. The performance of the designed is compared with the existing Chikungunya disease predicting models. General experimental study shows that the proposed ANFIS outperforms competitive models in all aspects, and F-score, accuracy, sensitivity, and specificity are found to be 97.5%, 97.3%, 97.1%, 97.14% respectively. Thus forecasting in terms of predicting Chikungunya disease using optimized machine learning model provides better results and good decision making.