Intuitionistic fuzzy time series forecasting method for non-stationary time series data with suitable number of clusters and different window size for fuzzy rule generation

The strict non-stationary time series (NS-TS) forecasting is one of the challenging tasks as the series does not follow any defined pattern. Previous studies had mainly focused on stationary, seasonal, or trending time series. This study aims to present an effective method for non-stationary time series (NS-TS) forecasting using the intuitionistic fuzzy time series clustering technique. The algorithm is proposed based on the observations and results obtained after the implementation of three existing algorithms with four variants of each. We have used four datasets to test and compare the performance of the proposed model. The experimental results suggest that the method can forecast the NS-TS effectively and more accurately as compared to existing methods.