Summary of Causalstock: Deep End-to-end Causal Discovery For News-driven Stock Movement Prediction, by Shuqi Li et al.
CausalStock: Deep End-to-end Causal Discovery for News-driven Stock Movement Prediction
by Shuqi Li, Yuebo Sun, Yuxin Lin, Xin Gao, Shuo Shang, Rui Yan
First submitted to arxiv on: 10 Nov 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Artificial Intelligence (cs.AI); Computational Engineering, Finance, and Science (cs.CE); Computation and Language (cs.CL)
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 CausalStock framework tackles two key issues in news-driven multi-stock movement prediction: relation discovery and noise in news data. The framework discovers temporal causal relations between stocks using a lag-dependent temporal causal discovery mechanism, which is then encapsulated in a Functional Causal Model to predict stock movements. A Denoised News Encoder is also introduced, leveraging large language models (LLMs) to extract useful information from massive news data. Experimental results demonstrate the effectiveness of CausalStock on six real-world datasets, outperforming strong baselines and providing a clear prediction mechanism with good explainability. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary CausalStock is a new way to predict how stock prices will move based on news. Right now, there are two big problems with this task: figuring out which stocks are related and dealing with noise in the news data. The researchers propose a new framework that uses causal relations between stocks to make predictions. They use a special kind of graph model to find these relationships and then use them to predict stock prices. They also develop a way to clean up noisy news data by using large language models to extract important information. The results show that CausalStock is better than other methods for predicting stock movements, and it can even provide explanations for its predictions. |
Keywords
* Artificial intelligence * Encoder
