ANALYSIS OF MULTIVARIATE GENERALIZED SPACE TIME AUTOREGRESSIVE (MULTIVARIATE GSTAR) MODEL

• Main Author: Lutfi Wijaya, Muhammad
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/54981
• Institution: Institut Teknologi Bandung

This thesis discusses the analysis of space-time model using the Multivariate Generalized Space Time Autoregressive (Multivariate GSTAR) model. Fore- casting using Multivariate GSTAR model is carried out on space-time data with more than one variable. Modeling with Multivariate GSTAR(1;1) is car- ried out by using distance inverse weight. Because the inverse distance weight matrix still univariate for each location and variable, it is necessary to mo- dify the matrix weights so that there is a relationship between variables and location. The prediction of the Multivariate GSTAR(1;1) model was applied to the COVID-19 data for two cases in three locations of sub-regions of Bandung City. Case I used positive, recovered, and died variables while Case II used close contacts, suspects, and positive variables. The prediction results of the Multivariate GSTAR(1;1) model in Case II were better than Case I.