{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,31]],"date-time":"2026-07-31T22:11:45Z","timestamp":1785535905691,"version":"3.56.0"},"reference-count":37,"publisher":"The Royal Society","issue":"2317","license":[{"start":{"date-parts":[[2026,4,9]],"date-time":"2026-04-09T00:00:00Z","timestamp":1775692800000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by\/4.0\/"},{"start":{"date-parts":[[2026,4,9]],"date-time":"2026-04-09T00:00:00Z","timestamp":1775692800000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"DOI":"10.13039\/501100000288","name":"Royal Society","doi-asserted-by":"publisher","id":[{"id":"10.13039\/501100000288","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2026,4,9]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:p>Symbolic regression (SR) is the machine learning (ML) method for learning functions from data. After a brief overview of the SR landscape, I will describe the two main challenges that traditional algorithms face: they have an unknown (and probably significant) probability of failing to find any given good function, and they suffer from ambiguity and poorly justified assumptions in their function-selection procedure. To address these, I propose an exhaustive search and model selection by the minimum description length (MDL) principle, which allows accuracy and complexity to be directly traded off by measuring each in units of information. I showcase the resulting publicly available Exhaustive Symbolic Regression (ESR) algorithm on three open problems in astrophysics: the expansion history of the universe, the effective behaviour of gravity in galaxies and the potential of the inflaton field. In each case, the algorithm identifies many functions superior to the literature standards. This general-purpose methodology should find widespread utility in science and beyond.<\/jats:p>\n                  <jats:p>This article is part of the discussion meeting issue \u2018Symbolic regression in the physical sciences\u2019.<\/jats:p>","DOI":"10.1098\/rsta.2024.0584","type":"journal-article","created":{"date-parts":[[2026,4,9]],"date-time":"2026-04-09T08:07:03Z","timestamp":1775722023000},"source":"Crossref","is-referenced-by-count":1,"title":["(Exhaustive) symbolic regression and model selection by minimum description length"],"prefix":"10.1098","volume":"384","author":[{"ORCID":"https:\/\/orcid.org\/0000-0003-0685-9791","authenticated-orcid":false,"given":"Harry","family":"Desmond","sequence":"first","affiliation":[{"name":"University of Portsmouth 1 , Portsmouth, UK","department":["Institute of Cosmology and Gravitation"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"175","published-online":{"date-parts":[[2026,4,9]]},"reference":[{"key":"2026051410263831400_R1","article-title":"syren-new: precise formulae for the linear and nonlinear matter power spectra with massive neutrinos and dynamical dark energy","author":"Sui","year":"2024","journal-title":"arXiv"},{"key":"2026051410263831400_R2","doi-asserted-by":"crossref","first-page":"433","DOI":"10.1093\/mind\/LIX.236.433","article-title":"I.\u2014Computing machinery and intelligence","volume":"LIX","author":"Turing","year":"1950","journal-title":"Mind"},{"key":"2026051410263831400_R3","volume-title":"Genetic algorithms in search, optimization and machine learning","author":"David","year":"1989"},{"key":"2026051410263831400_R4","volume-title":"Practical genetic algorithms","author":"Haupt","year":"2004","edition":"2nd edition"},{"key":"2026051410263831400_R5","doi-asserted-by":"crossref","unstructured":"Burlacu\n              B\n            , KronbergerG, KommendaM. 2020Operon c++: an efficient genetic programming framework for symbolic regression. InProceedings of the 2020 Genetic and Evolutionary Computation Conference Companion (GECCO 2020 Companion), pp. 1562\u20131570. New York, NY, USA: Association for Computing Machinery (ACM). (doi:10.1145\/3377929.3398099)","DOI":"10.1145\/3377929.3398099"},{"key":"2026051410263831400_R6","unstructured":"Cranmer\n              M\n            , Sanchez-GonzalezA, BattagliaP, XuR, CranmerK, SpergelD, HoS. 2020Discovering symbolic models from deep learning with inductive biases. InProceedings of the 2020 Genetic and Evolutionary Computation Conference Companion (GECCO 2020 Companion). New York, NY, USA: Association for Computing Machinery (ACM)."},{"key":"2026051410263831400_R7","unstructured":"Evolved Analytics LLC\n          . 2021Data Modeler 9.5.1. Evolved analytics LLC. Seehttps:\/\/evolved-analytics.com\/."},{"key":"2026051410263831400_R8","doi-asserted-by":"crossref","first-page":"333","DOI":"10.1109\/TEVC.2008.926486","article-title":"Order of nonlinearity as a complexity measure for models generated by symbolic regression via pareto genetic programming","volume":"13","author":"Vladislavleva","year":"2009","journal-title":"IEEE Trans. Evol. Comput."},{"key":"2026051410263831400_R9","doi-asserted-by":"crossref","first-page":"950","DOI":"10.1109\/TEVC.2023.3280250","article-title":"Exhaustive symbolic regression","volume":"28","author":"Bartlett","year":"2024","journal-title":"IEEE Trans. Evol. Computat."},{"key":"2026051410263831400_R10","article-title":"Exhaustive symbolic regression function sets","author":"Bartlett","year":"2022"},{"key":"2026051410263831400_R11","doi-asserted-by":"crossref","first-page":"563","DOI":"10.1086\/177173","article-title":"The structure of cold dark matter halos","volume":"462","author":"Navarro","year":"1996","journal-title":"APJ"},{"key":"2026051410263831400_R12","article-title":"Contemporary symbolic regression methods and their relative performance","author":"La Cava","year":"2021","journal-title":"arXiv"},{"key":"2026051410263831400_R13","first-page":"235","volume-title":"FFX: fast, scalable, deterministic symbolic regression technology","author":"McConaghy","year":"2011"},{"key":"2026051410263831400_R14","article-title":"An approach to symbolic regression using Feyn","author":"Ren\u00e9 Brol\u00f8s","year":"2021","journal-title":"arXiv"},{"key":"2026051410263831400_R15","article-title":"PySR: fast & parallelized symbolic regression in Python\/Julia","author":"Cranmer","year":"2020"},{"key":"2026051410263831400_R16","doi-asserted-by":"crossref","first-page":"465","DOI":"10.1016\/0005-1098(78)90005-5","article-title":"Modeling by shortest data description","volume":"14","author":"Rissanen","year":"1978","journal-title":"Automatica"},{"key":"2026051410263831400_R17","article-title":"Minimum description length revisited","author":"Gr\u00fcnwald","year":"2019","journal-title":"arXiv"},{"key":"2026051410263831400_R18","doi-asserted-by":"crossref","DOI":"10.1002\/047174882X","volume-title":"Elements of information theory","author":"Cover","year":"2005"},{"key":"2026051410263831400_R19","article-title":"Priors for symbolic regression","author":"Bartlett","year":"2023","journal-title":"arXiv"},{"key":"2026051410263831400_R20","doi-asserted-by":"crossref","first-page":"400","DOI":"10.1109\/TASSP.1987.1165125","article-title":"Estimation of probabilities from sparse data for the language model component of a speech recognizer","volume":"35","author":"Katz","year":"1987","journal-title":"IEEE Trans. Acoust. Speech Signal Process"},{"key":"2026051410263831400_R21","doi-asserted-by":"crossref","DOI":"10.1007\/s41114-022-00040-z","article-title":"Unveiling the Universe with emerging cosmological probes","volume":"25","author":"Moresco","year":"2022","journal-title":"Living Rev. Relativ."},{"key":"2026051410263831400_R22","article-title":"The Pantheon+ analysis: the full dataset and light-curve release","author":"Scolnic","year":"2021","journal-title":"arXiv"},{"key":"2026051410263831400_R23","doi-asserted-by":"crossref","first-page":"1817","DOI":"10.1093\/mnras\/stad597","article-title":"On the functional form of the radial acceleration relation","volume":"521","author":"Desmond","year":"2023","journal-title":"MNRAS"},{"key":"2026051410263831400_R24","doi-asserted-by":"crossref","first-page":"365","DOI":"10.1086\/161130","article-title":"A modification of the Newtonian dynamics as a possible alternative to the hidden mass hypothesis","volume":"270","author":"Milgrom","year":"1983","journal-title":"ApJ"},{"key":"2026051410263831400_R25","doi-asserted-by":"crossref","first-page":"384","DOI":"10.1086\/161132","article-title":"A modification of the Newtonian dynamics - implications for galaxy systems","volume":"270","author":"Milgrom","year":"1983","journal-title":"ApJ"},{"key":"2026051410263831400_R26","doi-asserted-by":"crossref","first-page":"371","DOI":"10.1086\/161131","article-title":"A modification of the Newtonian dynamics - implications for galaxies","volume":"270","author":"Milgrom","year":"1983","journal-title":"ApJ"},{"key":"2026051410263831400_R27","article-title":"Modified Newtonian dynamics: observational successes and failures","author":"Desmond","year":"2025","journal-title":"arXiv"},{"key":"2026051410263831400_R28","doi-asserted-by":"crossref","first-page":"152","DOI":"10.3847\/1538-4357\/836\/2\/152","article-title":"One law to rule them all: the radial acceleration relation of galaxies","volume":"836","author":"Lelli","year":"2017","journal-title":"ApJ"},{"key":"2026051410263831400_R29","doi-asserted-by":"crossref","first-page":"3342","DOI":"10.1093\/mnras\/stad2762","article-title":"The underlying radial acceleration relation","volume":"526","author":"Desmond","year":"2023","journal-title":"MNRAS"},{"key":"2026051410263831400_R30","doi-asserted-by":"crossref","first-page":"6130","DOI":"10.1093\/mnras\/stad2675","article-title":"On the fundamentality of the radial acceleration relation for late-type galaxy dynamics","volume":"525","author":"Stiskalek","year":"2023","journal-title":"MNRAS"},{"key":"2026051410263831400_R31","doi-asserted-by":"crossref","first-page":"020","DOI":"10.1088\/1475-7516\/2024\/04\/020","article-title":"Radial acceleration relation of galaxies with joint kinematic and weak-lensing data","volume":"2024","author":"Mistele","year":"2024","journal-title":"JCAP"},{"key":"2026051410263831400_R32","doi-asserted-by":"crossref","DOI":"10.1103\/PhysRevD.109.083524","article-title":"Optimal inflationary potentials","volume":"109","author":"Sousa","year":"2024","journal-title":"PRD"},{"key":"2026051410263831400_R33","first-page":"arXiv:1303.3787","article-title":"Encyclop\u00e6dia Inflationaris","author":"Martin","year":"2014","journal-title":"arXiv"},{"key":"2026051410263831400_R34","article-title":"Planck 2018 results. X. Constraints on inflation","volume":"641","author":"Planck Collaboration","year":"2020","journal-title":"AAP"},{"key":"2026051410263831400_R35","doi-asserted-by":"crossref","first-page":"062","DOI":"10.1088\/1475-7516\/2023\/04\/062","article-title":"Updated constraints on amplitude and tilt of the tensor primordial spectrum","volume":"2023","author":"Galloni","year":"2023","journal-title":"JCAP"},{"key":"2026051410263831400_R36","doi-asserted-by":"crossref","unstructured":"Burlacu\n              B\n            \n          . 2023GECCO\u20192022 Symbolic Regression Competition: Post-Analysis of the Operon Framework. InProc. of the Companion Conference on Genetic and Evolutionary Computation GECCO \u201923 Companion, Lisbon Portugal, pp. 2412\u20132419. New York, NY, USA: Association for Computing Machinery. (doi:10.1145\/3583133.3596390)","DOI":"10.1145\/3583133.3596390"},{"key":"2026051410263831400_R37","doi-asserted-by":"crossref","first-page":"297","DOI":"10.1086\/154079","article-title":"An analytic expression for the luminosity function for galaxies","volume":"203","author":"Schechter","year":"1976","journal-title":"APJ"}],"container-title":["Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/royalsocietypublishing.org\/rsta\/article-pdf\/doi\/10.1098\/rsta.2024.0584\/6131786\/rsta.2024.0584.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/royalsocietypublishing.org\/rsta\/article-pdf\/doi\/10.1098\/rsta.2024.0584\/6131786\/rsta.2024.0584.pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,5,14]],"date-time":"2026-05-14T14:42:10Z","timestamp":1778769730000},"score":1,"resource":{"primary":{"URL":"https:\/\/royalsocietypublishing.org\/rsta\/article\/384\/2317\/20240584\/481209\/Exhaustive-symbolic-regression-and-model-selection"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,4,9]]},"references-count":37,"journal-issue":{"issue":"2317","published-online":{"date-parts":[[2026,4,9]]},"published-print":{"date-parts":[[2026,4,9]]}},"URL":"https:\/\/doi.org\/10.1098\/rsta.2024.0584","relation":{},"ISSN":["1364-503X","1471-2962"],"issn-type":[{"value":"1364-503X","type":"print"},{"value":"1471-2962","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,4,9]]},"article-number":"20240584"}}