A computational framework for physics-informed symbolic regression with straightforward integration of domain knowledge
Abstract Discovering a meaningful symbolic expression that explains experimental data is a fundamental challenge in many scientific fields. We present a novel, open-source computational framework called Scientist-Machine Equation Detector (SciMED), which integrates scientific discipline wisdom in a...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2023-01-01
|
| Series: | Scientific Reports |
| Online Access: | https://doi.org/10.1038/s41598-023-28328-2 |
| _version_ | 1811175923082330112 |
|---|---|
| author | Liron Simon Keren Alex Liberzon Teddy Lazebnik |
| author_facet | Liron Simon Keren Alex Liberzon Teddy Lazebnik |
| author_sort | Liron Simon Keren |
| collection | DOAJ |
| description | Abstract Discovering a meaningful symbolic expression that explains experimental data is a fundamental challenge in many scientific fields. We present a novel, open-source computational framework called Scientist-Machine Equation Detector (SciMED), which integrates scientific discipline wisdom in a scientist-in-the-loop approach, with state-of-the-art symbolic regression (SR) methods. SciMED combines a wrapper selection method, that is based on a genetic algorithm, with automatic machine learning and two levels of SR methods. We test SciMED on five configurations of a settling sphere, with and without aerodynamic non-linear drag force, and with excessive noise in the measurements. We show that SciMED is sufficiently robust to discover the correct physically meaningful symbolic expressions from the data, and demonstrate how the integration of domain knowledge enhances its performance. Our results indicate better performance on these tasks than the state-of-the-art SR software packages , even in cases where no knowledge is integrated. Moreover, we demonstrate how SciMED can alert the user about possible missing features, unlike the majority of current SR systems. |
| first_indexed | 2024-04-10T19:43:50Z |
| format | Article |
| id | doaj.art-1a798436a5ae4f43ad7a7d55dd4901b7 |
| institution | Directory Open Access Journal |
| issn | 2045-2322 |
| language | English |
| last_indexed | 2024-04-10T19:43:50Z |
| publishDate | 2023-01-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj.art-1a798436a5ae4f43ad7a7d55dd4901b72023-01-29T12:08:41ZengNature PortfolioScientific Reports2045-23222023-01-0113111710.1038/s41598-023-28328-2A computational framework for physics-informed symbolic regression with straightforward integration of domain knowledgeLiron Simon Keren0Alex Liberzon1Teddy Lazebnik2Turbulence Structure Laboratory, School of Mechanical Engineering, Tel Aviv UniversityTurbulence Structure Laboratory, School of Mechanical Engineering, Tel Aviv UniversityDepartment of Cancer Biology, Cancer Institute, University College LondonAbstract Discovering a meaningful symbolic expression that explains experimental data is a fundamental challenge in many scientific fields. We present a novel, open-source computational framework called Scientist-Machine Equation Detector (SciMED), which integrates scientific discipline wisdom in a scientist-in-the-loop approach, with state-of-the-art symbolic regression (SR) methods. SciMED combines a wrapper selection method, that is based on a genetic algorithm, with automatic machine learning and two levels of SR methods. We test SciMED on five configurations of a settling sphere, with and without aerodynamic non-linear drag force, and with excessive noise in the measurements. We show that SciMED is sufficiently robust to discover the correct physically meaningful symbolic expressions from the data, and demonstrate how the integration of domain knowledge enhances its performance. Our results indicate better performance on these tasks than the state-of-the-art SR software packages , even in cases where no knowledge is integrated. Moreover, we demonstrate how SciMED can alert the user about possible missing features, unlike the majority of current SR systems.https://doi.org/10.1038/s41598-023-28328-2 |
| spellingShingle | Liron Simon Keren Alex Liberzon Teddy Lazebnik A computational framework for physics-informed symbolic regression with straightforward integration of domain knowledge Scientific Reports |
| title | A computational framework for physics-informed symbolic regression with straightforward integration of domain knowledge |
| title_full | A computational framework for physics-informed symbolic regression with straightforward integration of domain knowledge |
| title_fullStr | A computational framework for physics-informed symbolic regression with straightforward integration of domain knowledge |
| title_full_unstemmed | A computational framework for physics-informed symbolic regression with straightforward integration of domain knowledge |
| title_short | A computational framework for physics-informed symbolic regression with straightforward integration of domain knowledge |
| title_sort | computational framework for physics informed symbolic regression with straightforward integration of domain knowledge |
| url | https://doi.org/10.1038/s41598-023-28328-2 |
| work_keys_str_mv | AT lironsimonkeren acomputationalframeworkforphysicsinformedsymbolicregressionwithstraightforwardintegrationofdomainknowledge AT alexliberzon acomputationalframeworkforphysicsinformedsymbolicregressionwithstraightforwardintegrationofdomainknowledge AT teddylazebnik acomputationalframeworkforphysicsinformedsymbolicregressionwithstraightforwardintegrationofdomainknowledge AT lironsimonkeren computationalframeworkforphysicsinformedsymbolicregressionwithstraightforwardintegrationofdomainknowledge AT alexliberzon computationalframeworkforphysicsinformedsymbolicregressionwithstraightforwardintegrationofdomainknowledge AT teddylazebnik computationalframeworkforphysicsinformedsymbolicregressionwithstraightforwardintegrationofdomainknowledge |