Time-varying spatio-temporal autoregressive approach improves model flexibility in non-stationary data, suggesting enhanced accuracy in predictions.
The spatio-temporal autoregressive moving average (STARMA) model is frequently used in several studies concerning multivariate stationary time series. However, in practice, the assumption of stationarity is not always guaranteed. One way to proceed is to consider locally stationary processes. In this paper, we propose a time-varying spatio-temporal autoregressive and moving average (tvSTARMA) modelling based on the local stationarity assumption. The time-varying parameters are expanded as linear combinations of wavelet bases, and some procedures are proposed to estimate the coefficients. Some simulations and an application to historical weekly mean temperature records of Western states of the USA are illustrated.
No takes yet. Share an insight, caveat, or question.
Chen et al. (2025) studied this question.
Synapse has enriched 2 closely related papers on similar clinical questions. Consider them for comparative context: