Hybrid Fuzzy Autoregressive Integrated Moving Average Model with Adjusted Fuzzy Number

Abstract

 In recent years, a number of studies have demonstrated the efficiency, adaptability and accuracy of fuzzy time series forecasting models in the quest to improve the predictive power of the extant models. However, achievement of prediction accuracy is still one of the major challenges of these techniques. In this work, we proposed a model that combines the basic concepts of Fuzzy Autoregressive Integrated Moving Average (FARIMA) and Fuzzy Regression (FR) models which help to enhance prediction accuracy by narrowing down the projection bias problems specifically associated with the FARIMA model’s interval of possibility. The method is tested on real data obtained from the literature. In most of the instances, the experimental results indicated that the proposed method achieves a narrower interval of possibility compared to the existing methods in the literature.