Summary of D-pad: Deep-shallow Multi-frequency Patterns Disentangling For Time Series Forecasting, by Xiaobing Yuan and Ling Chen
D-PAD: Deep-Shallow Multi-Frequency Patterns Disentangling for Time Series Forecasting
by Xiaobing Yuan, Ling Chen
First submitted to arxiv on: 26 Mar 2024
Categories
- Main: Artificial Intelligence (cs.AI)
- Secondary: None
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The proposed D-PAD neural network disentangles intricate temporal patterns in time series forecasting by combining decomposition techniques with deep learning. The model consists of a multi-component decomposing block, a decomposition-reconstruction-decomposition module, and an interaction and fusion module. These components work together to extract frequency ranges and progressively extract information from mixed frequencies. Experimental results on seven real-world datasets demonstrate that D-PAD outperforms the best baseline by an average of 9.48% and 7.15% in MSE and MAE, respectively. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary D-PAD is a new way to predict future events based on past patterns. It uses a combination of old techniques (decomposition) with modern deep learning methods. This helps the model understand complex patterns like trends and seasons in time series data. The model has three main parts: decomposing, reconstructing, and combining. Tests were done on seven real-world datasets, and D-PAD did better than other models by a lot. |
Keywords
» Artificial intelligence » Deep learning » Mae » Mse » Neural network » Time series
