Fuzzy time series forecasting model based on second order fuzzy logical relationship and similarity measure / Nik Muhammad Farhan Hakim Nik Badrul Alam and Nazirah Ramli

Various fuzzy time series (FTS) forecasting methods have been proposed to cater for data in linguistic values. In this paper, an improved FTS forecasting method based on second order fuzzy logical relationship is proposed and it is used to forecast the enrollment of students in the University of Alabama. The performance of the forecasted results is compared to the actual data by using seven different similarity measures. The hybrid similarity measure based on geometric distance, centre of gravity, area, perimeter and height gives the best performance.