Mixed order single variable intuitionistic fuzzy time series forecasting method based on a new artificial neural network and grey wolf optimization algorithm

The ease of use of fuzzy time series and its success in forecasting performance has led to a rapid increase in this field. Although fuzzy time series methods work with membership values, intuitionistic fuzzy time series methods work based on both membership values and non-membership values. This study proposes a new mixed-order single variable intuitionistic fuzzy time series method for forecasting. The proposed method is based on a artificial neural network, intuitionistic fuzzy c-means algorithm and grey wolf optimization algorithm. The intuitionistic fuzzy time series is defined by using crisp values, memberships and non-memberships values. The fuzzy relations are determined based on a new artificial neural network based on the dendritic neuron model and grey wolf optimization algorithm. Forecast models will be created in two different ways based on membership and non-membership values, and the final forecasts will be obtained as a result of combining these models with the weights obtained by the grey wolf algorithm. The performance of the proposed method is compared with selected fuzzy methods in the literature by using different real-world time series.