Search: onr:"swepub:oai:DiVA.org:uu-461144" >
Deep multi-task min...
Deep multi-task mining Calabi-Yau four-folds
-
- Erbin, Harold (author)
- MIT, Ctr Theoret Phys, Cambridge, MA 02139 USA.;NSF Inst Artificial Intelligence & Fundamental In, Cambridge, MA 02139 USA.;Univ Paris Saclay, CEA, LIST, F-91120 Palaiseau, France.
-
- Finotello, Riccardo (author)
- Univ Paris Saclay, CEA, LIST, F-91120 Palaiseau, France.;Univ Paris Saclay, Serv Etud Analyt & Reactivite Surfaces SEAR, CEA, F-91191 Gif Sur Yvette, France.
-
- Schneider, Robin (author)
- Uppsala universitet,Teoretisk fysik
-
show more...
-
- Tamaazousti, Mohamed (author)
- Univ Paris Saclay, CEA, LIST, F-91120 Palaiseau, France.
-
show less...
-
MIT, Ctr Theoret Phys, Cambridge, MA 02139 USA;NSF Inst Artificial Intelligence & Fundamental In, Cambridge, MA 02139 USA.;Univ Paris Saclay, CEA, LIST, F-91120 Palaiseau, France. Univ Paris Saclay, CEA, LIST, F-91120 Palaiseau, France.;Univ Paris Saclay, Serv Etud Analyt & Reactivite Surfaces SEAR, CEA, F-91191 Gif Sur Yvette, France. (creator_code:org_t)
- 2021-11-25
- 2022
- English.
-
In: Machine Learning. - : IOP Publishing Ltd. - 2632-2153. ; 3:1
- Related links:
-
https://doi.org/10.1...
-
show more...
-
https://uu.diva-port... (primary) (Raw object)
-
https://doi.org/10.1...
-
https://urn.kb.se/re...
-
https://doi.org/10.1...
-
show less...
Abstract
Subject headings
Close
- We continue earlier efforts in computing the dimensions of tangent space cohomologies of Calabi-Yau manifolds using deep learning. In this paper, we consider the dataset of all Calabi-Yau four-folds constructed as complete intersections in products of projective spaces. Employing neural networks inspired by state-of-the-art computer vision architectures, we improve earlier benchmarks and demonstrate that all four non-trivial Hodge numbers can be learned at the same time using a multi-task architecture. With 30% (80%) training ratio, we reach an accuracy of 100% for h((1,1)) h((2,1)) (100% for both), 81% (96%) for h((3,1)), and 49% (83%) for h((2,2)). Assuming that the Euler number is known, as it is easy to compute, and taking into account the linear constraint arising from index computations, we get 100% total accuracy.
Subject headings
- NATURVETENSKAP -- Fysik -- Subatomär fysik (hsv//swe)
- NATURAL SCIENCES -- Physical Sciences -- Subatomic Physics (hsv//eng)
Keyword
- Calabi-Yau
- Hodge numbers
- string theory
- multi-task learning
- deep mining
- inception modules
- algebraic geometry
Publication and Content Type
- ref (subject category)
- art (subject category)
Find in a library
To the university's database