Using Clustering and Predictive Analysis of Infected Area on Dengue Outbreaks in Malaysia

Abstract

Machine learning and data mining have a great impact on the predictive analysis process. The features classification on machine learning can be used to adopt the clustering method to define further analysis on the targeted issues. Nowadays, the epidemic disease outbreaks have caused a great concern towards Malaysian community as the diseases can cause great fatality. One of the common killer epidemic diseases in Malaysia is dengue fever. Dengue fever is caused by dengue virus that spreads by Aedes mosquitoes. The outbreaks cause several cases of death and it varies throughout the states in Malaysia. The factors that cause this epidemic disease were determined and the data on the dengue outbreaks in Malaysia were gathered. To predict the infected area of dengue, data were mined and the machine learning method was implemented. In this study, the clustering method in machine learning for predictive analysis is proven to be an effective method in determining the most infected area of dengue outbreaks in Malaysia: Selangor and W.P. Kuala Lumpur/ Putrajaya. The selected areas were identified as the busiest place in Malaysia with a great number of population that had caused high physical contact and promoted the dengue outbreaks.