A new procedure of space-time modeling through the Invers of Autocovariance Matrix (IAcM) is proposed. By evaluating the IAcM behaviors on behalf of the Generalized Space-Time Autoregressive (GSTAR) process stationarity, we may find an appropriate model to space-time data series. This method can complete the Space-Time ACF and PACF methods for identifying space-time models. For study case, we apply the GSTAR models to the monthly tea production of some plantations in West Java, Indonesia.

s The movement of positive people Coronavirus Disease that was discovered in 2019 (Covid-19), written 2019-nCoV, from one location to another has a great opportunity to transmit the virus to more people. High-risk locations for transmission of the virus are public transportations, one of which is the train, because many people take turns in or together inside. One of the policies of the government is physical distancing, then followed by large-scale social restrictions. The keys to the policy are distance and movement. The most famous transportation used for the movement of people among provinces on Java is train. Here a Generalized Space Time Autoregressive (GSTAR) model is applied to forecast infected case of 2019-nCoV for 6 provinces in Java. The specialty of this model is the weight matrix as a tool to see spatial dependence. Here, the modified Inverse Distance Weight matrix is proposed as a combination of the population ratio factor with the average distance of an inter-provincial train on the island of Java. The GSTAR model (1; 1) can capture the pattern of daily cases increase in 2019-nCoV, evidenced by representative results, especially in East Java, where the increase in cases is strongly influenced by other provinces on the island of Java. Based on the Mean Squares of Residuals, it is obtained that the modified matrix gives better result in both estimating (in-sample) and forecasting (out-sample) compare with the ordinary matrix.

No abstract

Abstract. For GSTAR models, the least squares estimation method is commonly used since errors are assumed be uncorrelated. However, this method is not appropriate when errors are correlated, either in time or spatially. For these cases, the generalized least squares (GLS) method can be applied. GLS is more powerful since it has an error parameter that can act as a controller of the model to produce an efficient estimator. In this study, R Software was used to estimate GSTAR parameters. The resulted model was applied to real data, i.e. the monthly tea production of five plantations in West Java, Indonesia. The best model for forecasting was the GSTAR(1;1) model with temporally correlated error assumption.

No abstract