Modeling seasonality using the fuzzy integrated logical forecasting (FILF) approach

Fuzzy time series forecasting (FTSF) is a useful tool for forecasting without expert consultation as well as a user-friendly solution for non-expert forecasters. Before selecting the proper forecasting model, analysis of data series is a key step in the implementation of fuzzy time series forecasting. Seasonality is one of the change-making dimensions of data series that also include temperature, rainfall, freight rates and vessel traffic. The aim of this paper is to improve the fuzzy integrated logical forecasting (FILF) model for the seasonal time series by using the bivariate fuzzy time series approach. The proposed model is applied on the volume of vessel traffic on the Istanbul Strait in order to compare the accuracy of the proposed model with benchmark methods. In addition, the histogram damping partition (HDP) is used to define the initial length of intervals for the fuzzy C-means clustering method.