Advancing mathematics by guiding human intuition with AI
The practice of mathematics involves discovering patterns and using these to formulate and prove conjectures, resulting in theorems. Since the 1960s, mathematicians have used computers to assist in the discovery of patterns and formulation of conjectures1, most famously in the Birch and Swinnerton-D...
| Main Authors: | , , , , , , , , , , , , , |
|---|---|
| Format: | Journal article |
| Language: | English |
| Published: |
Nature Research
2021
|
| _version_ | 1797106544485072896 |
|---|---|
| author | Davies, A Velickovic, P Buesing, L Blackwell, S Zheng, D Tomasev, N Tanburn, R Battaglia, P Blundell, C Juhasz, A Lackenby, M Williamson, G Hassabis, D Kohli, D |
| author_facet | Davies, A Velickovic, P Buesing, L Blackwell, S Zheng, D Tomasev, N Tanburn, R Battaglia, P Blundell, C Juhasz, A Lackenby, M Williamson, G Hassabis, D Kohli, D |
| author_sort | Davies, A |
| collection | OXFORD |
| description | The practice of mathematics involves discovering patterns and using these to formulate and prove conjectures, resulting in theorems. Since the 1960s, mathematicians have used computers to assist in the discovery of patterns and formulation of conjectures1, most famously in the Birch and Swinnerton-Dyer conjecture2, a Millennium Prize Problem3. Here we provide examples of new fundamental results in pure mathematics that have been discovered with the assistance of machine learning—demonstrating a method by which machine learning can aid mathematicians in discovering new conjectures and theorems. We propose a process of using machine learning to discover potential patterns and relations between mathematical objects, understanding them with attribution techniques and using these observations to guide intuition and propose conjectures. We outline this machine-learning-guided framework and demonstrate its successful application to current research questions in distinct areas of pure mathematics, in each case showing how it led to meaningful mathematical contributions on important open problems: a new connection between the algebraic and geometric structure of knots, and a candidate algorithm predicted by the combinatorial invariance conjecture for symmetric groups4. Our work may serve as a model for collaboration between the fields of mathematics and artificial intelligence (AI) that can achieve surprising results by leveraging the respective strengths of mathematicians and machine learning.
|
| first_indexed | 2024-03-07T07:04:00Z |
| format | Journal article |
| id | oxford-uuid:f1e61139-46a9-46f7-ba7f-e85d115ed779 |
| institution | University of Oxford |
| language | English |
| last_indexed | 2024-03-07T07:04:00Z |
| publishDate | 2021 |
| publisher | Nature Research |
| record_format | dspace |
| spelling | oxford-uuid:f1e61139-46a9-46f7-ba7f-e85d115ed7792022-04-07T14:42:26ZAdvancing mathematics by guiding human intuition with AIJournal articlehttp://purl.org/coar/resource_type/c_dcae04bcuuid:f1e61139-46a9-46f7-ba7f-e85d115ed779EnglishSymplectic ElementsNature Research2021Davies, AVelickovic, PBuesing, LBlackwell, SZheng, DTomasev, NTanburn, RBattaglia, PBlundell, CJuhasz, ALackenby, MWilliamson, GHassabis, DKohli, DThe practice of mathematics involves discovering patterns and using these to formulate and prove conjectures, resulting in theorems. Since the 1960s, mathematicians have used computers to assist in the discovery of patterns and formulation of conjectures1, most famously in the Birch and Swinnerton-Dyer conjecture2, a Millennium Prize Problem3. Here we provide examples of new fundamental results in pure mathematics that have been discovered with the assistance of machine learning—demonstrating a method by which machine learning can aid mathematicians in discovering new conjectures and theorems. We propose a process of using machine learning to discover potential patterns and relations between mathematical objects, understanding them with attribution techniques and using these observations to guide intuition and propose conjectures. We outline this machine-learning-guided framework and demonstrate its successful application to current research questions in distinct areas of pure mathematics, in each case showing how it led to meaningful mathematical contributions on important open problems: a new connection between the algebraic and geometric structure of knots, and a candidate algorithm predicted by the combinatorial invariance conjecture for symmetric groups4. Our work may serve as a model for collaboration between the fields of mathematics and artificial intelligence (AI) that can achieve surprising results by leveraging the respective strengths of mathematicians and machine learning. |
| spellingShingle | Davies, A Velickovic, P Buesing, L Blackwell, S Zheng, D Tomasev, N Tanburn, R Battaglia, P Blundell, C Juhasz, A Lackenby, M Williamson, G Hassabis, D Kohli, D Advancing mathematics by guiding human intuition with AI |
| title | Advancing mathematics by guiding human intuition with AI |
| title_full | Advancing mathematics by guiding human intuition with AI |
| title_fullStr | Advancing mathematics by guiding human intuition with AI |
| title_full_unstemmed | Advancing mathematics by guiding human intuition with AI |
| title_short | Advancing mathematics by guiding human intuition with AI |
| title_sort | advancing mathematics by guiding human intuition with ai |
| work_keys_str_mv | AT daviesa advancingmathematicsbyguidinghumanintuitionwithai AT velickovicp advancingmathematicsbyguidinghumanintuitionwithai AT buesingl advancingmathematicsbyguidinghumanintuitionwithai AT blackwells advancingmathematicsbyguidinghumanintuitionwithai AT zhengd advancingmathematicsbyguidinghumanintuitionwithai AT tomasevn advancingmathematicsbyguidinghumanintuitionwithai AT tanburnr advancingmathematicsbyguidinghumanintuitionwithai AT battagliap advancingmathematicsbyguidinghumanintuitionwithai AT blundellc advancingmathematicsbyguidinghumanintuitionwithai AT juhasza advancingmathematicsbyguidinghumanintuitionwithai AT lackenbym advancingmathematicsbyguidinghumanintuitionwithai AT williamsong advancingmathematicsbyguidinghumanintuitionwithai AT hassabisd advancingmathematicsbyguidinghumanintuitionwithai AT kohlid advancingmathematicsbyguidinghumanintuitionwithai |