Deep Mind AI supports mathematical research
Nature Digest Vol.19 No. 3 |doi : 10.1038/ndigest.2022.220320
原文：Nature (2021-12-01) |doi: 10.1038/D41586-021-03593-1 | Deepmind's AI Helps UNTANGLE THE MATHEMATICS OF KNOTS
Davide Castelvecchi
Can you use machine learning in the mathematics field?The research team, who thought so, let AI learn a lot of data and searched for patterns that mathematicians missed.The new form of AI and humans, in which mathematicians prove it, may be useful in other research fields that handle large datasets.
A mathematical connection that humans have overlooked have been discovered for the first time by machine learning.The research team of Deepmind, a major artificial intelligence (Deepmind, London), has tagged with mathematicians and challenged the two different issues: knot and symmetry research.The method of using AI supports mathematicians to discover new patterns on both issues, and mathematicians were able to work on the conventional methods."I was very surprised to know how the machine learning tool is useful as an intuitive guide," said Mark Lackenby at Oxford University (UK), one of the mathematicians who participated in the research."I didn't expect my prejudice to be overturned."The research team reports this method in the December 2, 2021 issue (A).Davies et al. Nature600, 70-74; 2021）、大規模データセットを扱う他の数学分野にも恩恵をもたらすかもしれないとしている。
