Summary of Causalformer: An Interpretable Transformer For Temporal Causal Discovery, by Lingbai Kong et al.
CausalFormer: An Interpretable Transformer for Temporal Causal Discovery
by Lingbai Kong, Wengen Li, Hanchen Yang, Yichao Zhang, Jihong Guan, Shuigeng Zhou
First submitted to arxiv on: 24 Jun 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Methodology (stat.ME)
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 This paper proposes CausalFormer, an interpretable transformer-based causal discovery model for uncovering temporal causality within time series data. The method trains a deep learning model on a prediction task using a multi-kernel causal convolution that aggregates input time series along the temporal dimension while respecting priority constraints. A decomposition-based causality detector then interprets the trained model’s global structure to identify potential causal relations and construct a causal graph. CausalFormer outperforms state-of-the-art methods on synthetic, simulated, and real-world datasets. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Causal discovery is important for understanding how time series data changes over time. Researchers have developed ways to use deep learning models to find these relationships, but they only look at certain parts of the model. This paper proposes a new method called CausalFormer that uses transformers to learn about causality and then looks at all parts of the model to make it more accurate. They tested their method on different kinds of data and showed that it works well. |
Keywords
* Artificial intelligence * Deep learning * Time series * Transformer
