Probabilistic Regular Tree Priors for Scientific Symbolic Reasoning

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Schneider, Tim, Totounferoush, Amin, Nowak, Wolfgang, Staab, Steffen
Format: Preprint
Published: 2023
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909219974807552
author Schneider, Tim
Totounferoush, Amin
Nowak, Wolfgang
Staab, Steffen
author_facet Schneider, Tim
Totounferoush, Amin
Nowak, Wolfgang
Staab, Steffen
contents Symbolic Regression (SR) allows for the discovery of scientific equations from data. To limit the large search space of possible equations, prior knowledge has been expressed in terms of formal grammars that characterize subsets of arbitrary strings. However, there is a mismatch between context-free grammars required to express the set of syntactically correct equations, missing closure properties of the former, and a tree structure of the latter. Our contributions are to (i) compactly express experts' prior beliefs about which equations are more likely to be expected by probabilistic Regular Tree Expressions (pRTE), and (ii) adapt Bayesian inference to make such priors efficiently available for symbolic regression encoded as finite state machines. Our scientific case studies show its effectiveness in soil science to find sorption isotherms and for modeling hyper-elastic materials.
format Preprint
id arxiv_https___arxiv_org_abs_2306_08506
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Probabilistic Regular Tree Priors for Scientific Symbolic Reasoning
Schneider, Tim
Totounferoush, Amin
Nowak, Wolfgang
Staab, Steffen
Machine Learning
Artificial Intelligence
Formal Languages and Automata Theory
Symbolic Regression (SR) allows for the discovery of scientific equations from data. To limit the large search space of possible equations, prior knowledge has been expressed in terms of formal grammars that characterize subsets of arbitrary strings. However, there is a mismatch between context-free grammars required to express the set of syntactically correct equations, missing closure properties of the former, and a tree structure of the latter. Our contributions are to (i) compactly express experts' prior beliefs about which equations are more likely to be expected by probabilistic Regular Tree Expressions (pRTE), and (ii) adapt Bayesian inference to make such priors efficiently available for symbolic regression encoded as finite state machines. Our scientific case studies show its effectiveness in soil science to find sorption isotherms and for modeling hyper-elastic materials.
title Probabilistic Regular Tree Priors for Scientific Symbolic Reasoning
topic Machine Learning
Artificial Intelligence
Formal Languages and Automata Theory
url https://arxiv.org/abs/2306.08506