Intuitionistic fuzzy set-based time series forecasting model via delegeration of hesitancy degree to the major grade de-i-fuzzification and arithmetic rules based on centroid defuzzification

Abstract

De-i-fuzzification is a process of converting the intuitionistic fuzzy set into a fuzzy set. It becomes one of the core procedures in fuzzy time series forecasting model based on the intuitionistic fuzzy set. In this paper, we propose a fuzzy time series forecasting model based on intuitionistic fuzzy set via de-i-fuzzification. The de-i-fuzzification approach used is assigning the hesitancy degree to the major grade. The data are partitioned into a few intervals using the frequency density-based method. The data in the fuzzy set form is then transformed into an intuitionistic fuzzy set using the definition of intuitionistic fuzzy set. The arithmetic rules based on centroid defuzzification is used to obtain the forecasted output. The model is implemented on the data of student enrolment at the University of Alabama. The results are then compared to forecasting method using classical fuzzy set and similar de-i-fuzzification approach using max-min operation. The proposed method outperforms the other two methods, thus supports the fact that intuitionistic fuzzy set is a generalization of a classical fuzzy set and gives better performance in forecasting.

Probabilistic fuzzy set and particle swarm optimization based computational method for fuzzy time series forecasting

“…Many IFS based FTS forecasting methods [17][18][19][20] were also developed using computational algorithm, pi-sigma artificial neural networks with artificial bee colony, long short -term memory, and function approach. Alam et al [21,22] used simple arithmetic rule for intuitionistic FTS forecasting. Recently, Yolcu & Yolcu [23] proposed a novel intuitionistic fuzzy time series forecasting model with cascaded structure for financial time series.…”

“…Bisht & Kumar [18] and Bisht & Kumar [40] devised computational approach based FTS forecasting methods using IFS and hesitant fuzzy set. Alam et al [21,22] used simple arithmetic rule for intuitionistic FTS forecasting by delegating hesitancy degree to the major grade intuitionistic defuzzification. Recently, Pant and Kumar [41] developed computational method for FTS forecasting using IFS and self organized direction aware algorithm.…”