Optimization the parameter of forecasting algorithm by using the genetical algorithm toward the information systems of geography for predicting the patient of dengue fever in district of sragen, Indonesia

Abstract

The major increase in number of the patients of Dengue Fever in district of Sragen, Indonesia in the last three years (2013–2015) has taken casualties in district of Sragen, Indonesia. It affected all the people in all ages including teenagers under 15 years old. The lacking of anticipation from Health Department of District Sragen was the result of unavailable system that could predict the increasing numbers of Dengue Fever's patients. In order to solve those issues, the author then did the sustainable research by applying the combination of Genetical Algorithm (GA) and Triple Exponential Smoothing (TES) to predict the patients of Dengue Fever in that district specially on years 2016 and will be continued. The data used by the researcher was the data of Dengue Fever's patient from 2013 to 2016 in the first semester. GA was used to cover the weakness part of TES in setting the parameter of alpha, beta, and gamma that influenced the accuracy of prediction. The result of this research was the comparative data between GA-TES and TES and also the calculation of the increasing accuracy after using GA. The calculation of accuracy itself used the method of Mean Absolute Percentage Error (MAPE). The data of testing result showed the average of increasing of the combination of GA-TES' algorithm which was 8% comparing to TES' algorithm. Contribution this research is in method repairing of econometric forecasting using optimization algorithm, hope there's improving new model of forecasting method weak which the accuracy increase to 0%.