Modeling autoregressive fuzzy time series data based on semi-parametricmethods

چکیده مقاله

In time series analysis, such as other statistical problems, we may confront imprecise quantity. One case is a situation in which the observations related to underlying systems are imprecise. This paper proposes a semi-parametric autoregressive model for those real-world applications whose observed data are reported by fuzzy numbers. To this end, a hybrid method including nonparametric kernel-based approach and the least absolute deviations is suggestedwhich allows us to estimate the parameters of the model and the fuzzy nonlinear function of the innovations, simultaneously. In order to examine the performance and effectiveness of the proposed fuzzy semi-parametric time series model, some common goodness-of-fit criteria are employed. The obtained results based on a practical example of simulated fuzzy time series data indicated that the proposed method is potentially effective for predicting fuzzy time series data.