Dynamic Type-2 Fuzzy Time Warping (DT2FTW): A Hybrid Model for Uncertain Time-Series Prediction

Abstract

Prediction of time series is associated with nondeterministic pattern analysis for uncertainconditions. Therefore, it is necessary to develop high-quality prediction methods for real-worldapplications. Type-2 fuzzy systems can handle high-order uncertainties, such as sequentialdependencies associated with time series. Precise and reliable prediction can help to developreasonable strategies and assist specialists in planning the best policies for modeling eventsin uncertain time series. In this study, a hybrid model (dynamic type-2 fuzzy time warping[DT2FTW]) was proposed for handling high-order uncertainties in time-series prediction. Atype-2 fuzzy intelligent system was developed alongside a dynamic time warping algorithmfor predicting the patterns’ similarity in long-time series for time-series prediction. The resultsdemonstrate that the proposed DT2FTW model yields more reliable predictions on globalstandard benchmarks such as the Mackey-Glass, Dow Jones, and NASDAQ time-series. Theresults also confirm that the proposed DT2FTW model has lower error rates than its counterpartalgorithms in terms of the root mean square error (RMSE), mean absolute error (MAE), andmean percentage error (MPE). In addition, the results confirm the superiority of the proposedmodel with an average area under the ROC curve (AUC) of 94%, with the 95% confidenceinterval (92%-95%).