Smoothing for Optimizing Arima and ANN Models for Predicting the Number of Treatments for TB Case Patients"

Abstract

Statistical models can be used to characterize numerical data to understand its behavior and patterns. For example, The TB care model can signal to local governments regarding when they should carry out prevention and prepare health care facilities. To find a model that can be optimized to predict the number of TB patient care occupancies so that the model has adequate performance. This research examines the performance of Wavelet-ARIMA-GARCH using tuning parameters in modeling and forecasting. Using Semarang City TB Incidence Treatment data from 2019 to 2022, this research concludes that the hybrid ANN -Wavelet-GARCH(1,1) model with parameter tuning is the best performance model.