A new explainable robust high-order intuitionistic fuzzy time-series method

Abstract

Fuzzy time series forecasting methods based on type-1 fuzzy sets continue to have largely proposed in the literature. These methods use only membership values in determining fuzzy relations. However, Intuitionistic fuzzy time series models basically use both membership values and nonmembership values. So, it can be considered that the using of intuitionistic fuzzy time forecasting models will be able to increase the forecasting performance in the fuzzy time series analyses because of the fact that more information is used. Therefore, Intuitionistic fuzzy time series models have started to use for solving the real-life series in the fuzzy time series literature since 2013. In this study, a new explainable robust high order intuitionistic fuzzy time series forecasting method are proposed based on new defined model. In the proposed method, the algorithm of intuitionistic fuzzy c-means is used for fuzzification of observations and a robust regression method is used to determine fuzzy relations. Because the robust regression is employed to define fuzzy relation, all inputs of the method can be explainable and they can be statistically tested and commented. Applications of this study have been made by using energy data of Primary Energy Consumption (PEC) between 1965 and 2016 for 23 countries in the region of Europe&Eurasia. Forecasting performances obtained from these applications by using the proposed method have been compared with performances of some other time series method in the literature and the results have been discussed.

Implementing lee's model to apply fuzzy time series in forecasting bitcoin price

“…Lee's model in constructing FLRG considers all relations interconnected and must be calculated because it affects the predictive value [20]. The following are the steps that must be passed in working on Lee's FTS method [21].…”

“…The middle value of each interval in the FLRG created in the last step was used to get the forecasted value. Defuzzification will replace the fuzzy output with a firm value based on the membership function to produce forecasting results [21].…”