Journals Proceedings

Abstract

Dengue Fever is one of acute and deadly diseases that commonly happens in tropical area. The spread of it is also influenced by geographical condition. Indonesia, Particularly in Malang that is a tropical area with a geographical condition supports the development of this disease. It needs a fat-moving action to the early step precaution so that the number of patients can be reduced. As the primary decision for an early prevention, it needs predictions about several cases of dengue fever of some period in the future. The result of this prediction is needed by Public Health Office of Malang as one of instances that responsible of dengue fever cases. This research analyses performance as a prediction model in getting the predictions in a number of dengue fever cases in Malang, Indonesia for some different group of data. The models suggested are Multiplicative Holt-Winters, Additive Holt-Winters, Multiplicative Decomposition and Autoregressive Moving Average (ARIMA). Those models are applied in special data to some cases in Malang which is categorized in 3 groups, namely Lowlands (Malang Rendah), Mediumlands (Malang Sedang), Highlands (Malang Tinggi). The result shows that Multiplicative Holt-Winters model is the best model for lowlands, and mediumlands. Meanwhile