Summary of Deep Generative Symbolic Regression, by Samuel Holt et al.
Deep Generative Symbolic Regression
by Samuel Holt, Zhaozhi Qian, Mihaela van der Schaar
First submitted to arxiv on: 30 Dec 2023
Categories
- Main: Machine Learning (cs.LG)
- 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 paper proposes a novel approach to symbolic regression (SR), which aims to discover concise closed-form mathematical equations from data. The existing methods for SR fail to scale with the number of input variables, making it a challenging task. To address this issue, the authors leverage pre-trained deep generative models to capture the intrinsic regularities of equations, providing a solid foundation for subsequent optimization steps. The proposed framework, Deep Generative Symbolic Regression (DGSR), unifies several prominent approaches of SR and offers a new perspective to justify and improve on previous ad hoc designs. Experimental results show that DGSR achieves a higher recovery rate of true equations in the setting of a larger number of input variables and is more computationally efficient at inference time than state-of-the-art RL symbolic regression solutions. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Symbolic regression tries to find simple math formulas from data, which is important for scientific discoveries. However, this task is very hard because the search space is huge. Previous methods didn’t scale well with the number of input variables. The authors thought that maybe there are some underlying rules or patterns in these formulas that could help us find them more efficiently. They use deep learning models to capture these patterns and then use those patterns to help find the formulas. This new approach, called DGSR, is better than previous methods at finding true equations when there are many input variables and is also faster. |
Keywords
* Artificial intelligence * Deep learning * Inference * Optimization * Regression
