A Comparison Between Backpropagation, Holt-Winter, and Polynomial Regression Methods in Forecasting Dog Bites Cases in Bali

Abstract

Rabies is a zoonotic disease that is usually transmitted to humans through animal bites. It can cause severe damage to the central nervous system and is generally fatal. Dog bite cases are considered the leading cause of rabies transmission in Bali. The government's preventive action is expected to reduce the problem of increasing the number of dog bite cases so that it does not spread quickly and cause casualties. Data mining is an attempt to extract knowledge from a set of data. The use of data mining in this study is to forecast the number of dog bite cases in Bali. Forecasting predicts what will happen in the future based on relevant data in the past and placing it in a mathematical model. Data mining methods used in forecasting dog bite cases are backpropagation, holt-winters, polynomial regression methods. This forecasting aims to help the government predict dog bite cases in the coming period to prepare appropriate countermeasures. Forecasting is done using data on bite cases every month in Bali province for five years, from 2015 to 2019. Dog bite case data is divided into four datasets for each attribute, namely data on the number of dog bite cases, the number of vaccinations, the number of male deaths, the number of female deaths. The four datasets are divided into training data and testing data to share 80% training data and 20% testing data. The results obtained are that the backpropagation method is better at predicting dog bite case data with an average MAPE error rate of 11.59%, while the holt-winters method has an average MAPE error rate of 16.05%, and the polynomial regression method has an average MAPE error rate of 19.91%

Keywords : Dog Bites, Rabies, Forecasting, Backpropagation, Holt-Winter, Polynomial Regression