Reproduce “Class Symbolic Regression: Gotta Fit 'Em All” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Wassim Tenachi, Rodrigo Ibata, Thibaut L. François, Foivos I. Diakogiannis
Researchers developed Class Symbolic Regression (Class SR), a hierarchical framework designed to discover a single analytical functional form that simultaneously fits multiple related datasets. The framework robustly identified underlying physical laws, outperforming traditional symbolic regression on synthetic benchmarks and accurately recovering a galactic potential from simulated stellar streams, especially under noisy conditions.
Key insights
- Leveraging information across multiple related datasets enhances the robustness and generalizability of discovered symbolic expressions, making them less prone to overfitting noise in individual observations.
- Hierarchical parameterization allows for the simultaneous identification of universal physical laws and their specific manifestations or boundary conditions across different realizations.
- The integration of deep reinforcement learning with an aggregated cost function across datasets enables an efficient search for a single, universal analytical form that optimally fits an entire class of phenomena.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.