Interval type-2 fuzzy C-means forecasting model for fuzzy time series

Abstract

The fuzzy time-series (FTS) prediction model is based on fuzzy theory, which handles uncertain and fuzzy data. Owing to its good interpretability and prediction accuracy, the model based on the fuzzy C-means (FCM) clustering algorithm is most widely used. However, the fuzzy parameter m in the FCM algorithm is an empirical value that leads to uncertainty in classification results, which can negatively affect predictions. This study proposes a new model, IT2-FCM-FTS, that uses the interval type-2 (IT2) FCM algorithm instead of the traditional FCM to divide the sample domain and improve the performance of the FTS model. To verify the reliability of the proposed model, five datasets and four evaluation indices are used to compare the prediction results of the proposed model with the traditional ARIMA model and the one based on FCM. The results show that the proposed model is superior to both in terms of prediction accuracy.