Cat swarm optimization and Levenberg-Marquardt for model identification and prediction identification and prediction tuberculosis disease spreading

Abstract

Tuberculosis (TB) is an infectious disease caused by the bacterium Mycobacterium tuberculosis and can cause death. Based on the prevalence rate, Indonesia ranks third after India and China as the countries with the highest number of TB sufferers in the world. Models identification and prediction are needed to minimize and anticipate problems that can occur. In applying the model, Cat Swarm Optimization and the Levenberg-Marquardt are used. The purpose of this study is to obtain the prediction result based on the model identification in TB disease spreading using a swarm optimization algorithm and Levenberg-Marquardt. The identification is intended to analyze the spread of TB disease based on actual data. The process begins with estimating the parameters in the model using the Cat Swarm Optimization algorithm. After getting the optimal parameters, the acquisition of the model uses Levenberg-Marquardt. Based on the implementation of the program on the spread of TB disease in East Java in the period of the first quarter of 2002 to the fourth quarter of 2017, the MSE of training and prediction process were 0.000225563 and 0.0085307, respectively. The MSE value from the prediction process shows that the Cat Swarm Optimization and Levenberg-Marquardt can be used to identify models and predict the spread of tuberculosis.