Summary of Drformer: Multi-scale Transformer Utilizing Diverse Receptive Fields For Long Time-series Forecasting, by Ruixin Ding et al.
DRFormer: Multi-Scale Transformer Utilizing Diverse Receptive Fields for Long Time-Series Forecasting
by Ruixin Ding, Yuqi Chen, Yu-Ting Lan, Wei Zhang
First submitted to arxiv on: 5 Aug 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Artificial Intelligence (cs.AI); Machine Learning (stat.ML)
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 A novel approach to long-term time series forecasting (LTSF) is proposed, utilizing a dynamic tokenizer and multi-scale Transformer model to capture diverse receptive fields and sparse patterns in time series data. The authors develop a dynamic sparse learning algorithm to segment data into sub-level patches, serving as input tokens for the patch-based transformer. Additionally, they introduce group-aware rotary position encoding to enhance intra- and inter-group position awareness among representations across different temporal scales. The proposed model, DRFormer, is evaluated on various real-world datasets, demonstrating its superiority compared to existing methods. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary A new way of predicting future events based on past data is being developed. This approach uses a special type of artificial intelligence called transformers and applies it to time series forecasting. Time series forecasting tries to predict what will happen in the future based on patterns we can see in the past. The new approach is better than existing methods at capturing different types of patterns that can occur in time series data, like trends or fluctuations. |
Keywords
» Artificial intelligence » Time series » Tokenizer » Transformer
