Simulation of Dengue Outbreak Prediction

Abstract

Neural Network Model (NNM), Hidden Markov Model (HMM) and Regression Model (RM) are developed to predict the spread of dengue outbreak in Malaysia. The case study covered dengue cases data from Selangor, which include seven mukims and eight administrative districts in year of 2004 and 2005. Specific criteria concerned are location, time (weeks) and intensity of dengue cases. Critical discussion of some previous studies upon the performance of each approach reveals that NNM has several advantages over the two other models in prediction although some limitations are observed and this indicated that NNM might have the better predict of dengue outbreak. However, these models will be further studied by measuring their Root Mean Square Error (RMSE) to identify the best prediction model.