Fuzzy Time Series Inference for Stationary Linear Processes: Features and Algorithms With Simulation

The primary objective of this article is to estimate the unknown parameters of stationary linear processes based on a fuzzy time series approach to observations that follow AR (1) processes. Predicted observations are obtained using fuzzy time series. Both actual and forecasted observations are utilized to study various classic method’s estimators for the autoregressive parameter. The comparisons between actual and forecasted observations in all estimating processes are discussed based on the mean squared errors. Furthermore, to investigate the extent to which fuzzy time series can enhance estimates produced by traditional estimating techniques. Based on these comparisons, it is possible to explore how fuzzy time series contribute to the improvement of classical methods’ estimations.