Interactive symbolic regression with co-design mechanism through offline reinforcement learning

Symbolic Regression holds great potential for uncovering underlying mathematical and physical relationships from observed data. However, the vast combinatorial space of possible expressions poses significant challenges for previous online search methods and pre-trained transformer models, which most...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:Nature communications Ročník 16; číslo 1; s. 3930 - 12
Hlavní autoři: Tian, Yuan, Zhou, Wenqi, Viscione, Michele, Dong, Hao, Kammer, David S., Fink, Olga
Médium: Journal Article
Jazyk:angličtina
Vydáno: London Nature Publishing Group UK 26.04.2025
Nature Publishing Group
Nature Portfolio
Témata:
639/705/1042
639/705/117
Autoregressive models
Benchmarks
Co-design
Combinatorial analysis
Humanities and Social Sciences
Learning
Mathematical models
multidisciplinary
Regression
Regression analysis
Reinforcement
Science
Science (multidisciplinary)
Subject specialists
ISSN:2041-1723, 2041-1723
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:Symbolic Regression holds great potential for uncovering underlying mathematical and physical relationships from observed data. However, the vast combinatorial space of possible expressions poses significant challenges for previous online search methods and pre-trained transformer models, which mostly do not consider the integration of domain experts’ prior knowledge. To address these challenges, we propose the Symbolic Q-network, an advanced interactive framework for large-scale symbolic regression. Unlike previous transformer-based SR approaches, Symbolic Q-network leverages reinforcement learning without relying on a transformer-based decoder. Furthermore, we propose a co-design mechanism, where the Symbolic Q-network facilitates effective interaction with domain experts at any stage of the equation discovery process. Our extensive experiments demonstrate Sym-Q performs comparably to existing pretrained models across multiple benchmarks. Furthermore, our experiments on real-world cases demonstrate that the interactive co-design mechanism significantly enhances Symbolic Q-network’s performance, achieving greater performance gains than standard autoregressive models. Symbolic Regression is a method for uncovering mathematical and physical relationships from data. Here, the authors propose Sym-Q, an interactive reinforcement learning-based framework with a co-design mechanism that enables efficient and guided equation discovery.
Bibliografie:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
content type line 23
ISSN:2041-1723
2041-1723
DOI:10.1038/s41467-025-59288-y