Learning Nonstationary Time-Series With Dynamic Pattern Extractions

The era of information explosion had prompted the accumulation of a tremendous amount of time-series data, including stationary and nonstationary time-series data. State-of-the-art algorithms have achieved a decent performance in dealing with stationary temporal data. However, traditional algorithms that tackle stationary time-series do not apply to nonstationary series like Forex trading. This article investigates applicable models that can improve the accuracy of forecasting future trends of nonstationary time-series sequences. In particular, we focus on identifying potential models and investigate the effects of recognizing patterns from historical data. We propose a combination of the seq2seq model based on recurrent neural network, along with an attention mechanism and an enriched set features extracted via dynamic time warping and zigzag peak valley indicators. Customized loss functions and evaluating metrics have been designed to focus more on the predicting sequence’s peaks and valley points. Our results show that our model can predict 4-h future trends with high accuracy in the Forex dataset, which is crucial in realistic scenarios to assist foreign exchange trading decision making. We further provide evaluations of the effects of various loss functions, evaluation metrics, model variants, and components on model performance.