Prediksi Jumlah Kasus Malaria di Labuhanbatu Utara Menggunakan Metode Gaussian Process Regression

Abstract

Malaria is a life threatening disease that can be found in tropical countries. Malaria is caused by the bite of an infected female Anopheles mosquito. Malaria is widely found in tropical countries, including Indonesia. In North Sumatra, some areas such as Asahan, Batubara, and Labuhanbatu Utara still have moderate endemic status for malaria. This study aims to predict the number of malaria cases in Labuhanbatu Utara using the Gaussian Process Regression (GPR) method as a countermeasure to the development of the number of malaria cases. The results of this study are expected to be able to predict the number of malaria cases to mitigate the development of malaria, estimate earlier the period at risk of experiencing a spike in cases, and become a reference for further research in predicting events using the GPR method. The data used includes weekly time, maximum temperature, sunshine duration, rainfall index, rainfall index, flood and inundation index, population density and growth index, and total malaria cases . In order to obtain a larger amount of data to improve model performance, this study used the Time Series Generative Adversarial Network (TGAN) method. This study shows that the GPR method with a combination of Matern and Dot Product kernels gets fairly accurate prediction results on malaria case data. It was found that models trained and tested using synthesized data were able to provide relatively accurate predictions and close to actual values on a smaller scale. This is shown by t he MSE evaluation metric value of 6.52, RMSE of 2.55 and MAE of 1.71. While the data trained and tested with the original data gives an evaluation metric value of MSE of 8.65, RMSE of 2.94 and MAE of 2.28. This shows that GPR is more suitable for use with a larger amount of data so that it can improve prediction accuracy.