D-Transformer: A Deep Learning Model for Time Series Prediction
Huyu Wu, Ruoyu Li, Gaojie Sheng, Daniel J. Wilson · 2023
The Transformer model is a deep learning model that has been extensively applied across diverse industries since its inception. However, the lack of Transformer models tailored for timing forecasting has led to the development of the proposed D-Transformer model in this study. The D-Transformer model is based on the Duffing equation and enables accurate forecasting of various types of time series data. The D-Transformer model is a type of transformer model that uses a modern and evolving mathematical techniques set, generally known as attention or self-attention. To evaluate the effectiveness of D-Transformer, a comparison was made with the conventional Transformer model. The results demonstrate that D-Transformer exhibits optimal prediction performance on power data sets, thereby validating the superiority of the proposed model over existing methods. In conclusion, the proposed D-Transformer model provides a promising solution for timing forecasting in various industries. Further investigations are warranted to assess the applicability and effectiveness of the model in other domains.