Categorical representation learning and RG flow operators for algorithmic classifiers
Following the earlier formalism of the categorical representation learning, we discuss the construction of the ‘RG-flow-based categorifier’. Borrowing ideas from the theory of renormalization group (RG) flows in quantum field theory, holographic duality, and hyperbolic geometry and combining them wi...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IOP Publishing
2023-01-01
|
| Series: | Machine Learning: Science and Technology |
| Subjects: | |
| Online Access: | https://doi.org/10.1088/2632-2153/acb488 |
| _version_ | 1827766702673231872 |
|---|---|
| author | Artan Sheshmani Yi-Zhuang You Wenbo Fu Ahmadreza Azizi |
| author_facet | Artan Sheshmani Yi-Zhuang You Wenbo Fu Ahmadreza Azizi |
| author_sort | Artan Sheshmani |
| collection | DOAJ |
| description | Following the earlier formalism of the categorical representation learning, we discuss the construction of the ‘RG-flow-based categorifier’. Borrowing ideas from the theory of renormalization group (RG) flows in quantum field theory, holographic duality, and hyperbolic geometry and combining them with neural ordinary differential equation techniques, we construct a new algorithmic natural language processing architecture, called the RG-flow categorifier or for short the RG categorifier, which is capable of data classification and generation in all layers. We apply our algorithmic platform to biomedical data sets and show its performance in the field of sequence-to-function mapping. In particular, we apply the RG categorifier to particular genomic sequences of flu viruses and show how our technology is capable of extracting the information from given genomic sequences, finding their hidden symmetries and dominant features, classifying them, and using the trained data to make a stochastic prediction of new plausible generated sequences associated with a new set of viruses which could avoid the human immune system. |
| first_indexed | 2024-03-11T11:45:41Z |
| format | Article |
| id | doaj.art-889f6b4a9eac4d0cbe7ba89e3b8e2113 |
| institution | Directory Open Access Journal |
| issn | 2632-2153 |
| language | English |
| last_indexed | 2024-03-11T11:45:41Z |
| publishDate | 2023-01-01 |
| publisher | IOP Publishing |
| record_format | Article |
| series | Machine Learning: Science and Technology |
| spelling | doaj.art-889f6b4a9eac4d0cbe7ba89e3b8e21132023-11-09T13:56:32ZengIOP PublishingMachine Learning: Science and Technology2632-21532023-01-014101501210.1088/2632-2153/acb488Categorical representation learning and RG flow operators for algorithmic classifiersArtan Sheshmani0Yi-Zhuang You1https://orcid.org/0000-0003-4080-5340Wenbo Fu2Ahmadreza Azizi3Department of Mathematics, and Harvard University Physics department, Center for Mathematical Sciences and Applications, Harvard University , Jefferson Laboratory, 17 Oxford St, Cambridge, MA, 02138, United States of America; Institute of the Mathematical Sciences of the Americas, University of Miami , 1365 Memorial Drive Ungar 515, Coral Gables, FL, 33146, United States of America; National Research University Higher School of Economics , Russian Federation, Laboratory of Mirror Symmetry, NRU HSE, 6 Usacheva str., Moscow, 119048, Russia; NSF AI Institute for Artificial Intelligence and Fundamental InteractionsDepartment of Physics, Condensed matter group, UC San Diego, University of California San Diego , 9500 Gilman Dr, La Jolla, CA, 92093, United States of AmericaQGNai INC. (Quantum Geometric networks for artificial intelligence) , 83 Cambridge Parkway, Unit W806, Cambridge, MA, 02142, United States of AmericaQGNai INC. (Quantum Geometric networks for artificial intelligence) , 83 Cambridge Parkway, Unit W806, Cambridge, MA, 02142, United States of AmericaFollowing the earlier formalism of the categorical representation learning, we discuss the construction of the ‘RG-flow-based categorifier’. Borrowing ideas from the theory of renormalization group (RG) flows in quantum field theory, holographic duality, and hyperbolic geometry and combining them with neural ordinary differential equation techniques, we construct a new algorithmic natural language processing architecture, called the RG-flow categorifier or for short the RG categorifier, which is capable of data classification and generation in all layers. We apply our algorithmic platform to biomedical data sets and show its performance in the field of sequence-to-function mapping. In particular, we apply the RG categorifier to particular genomic sequences of flu viruses and show how our technology is capable of extracting the information from given genomic sequences, finding their hidden symmetries and dominant features, classifying them, and using the trained data to make a stochastic prediction of new plausible generated sequences associated with a new set of viruses which could avoid the human immune system.https://doi.org/10.1088/2632-2153/acb488renormalization group flowneural ODEhyperbolic geometryholographic dualitycategory theorycategorical representation learning |
| spellingShingle | Artan Sheshmani Yi-Zhuang You Wenbo Fu Ahmadreza Azizi Categorical representation learning and RG flow operators for algorithmic classifiers Machine Learning: Science and Technology renormalization group flow neural ODE hyperbolic geometry holographic duality category theory categorical representation learning |
| title | Categorical representation learning and RG flow operators for algorithmic classifiers |
| title_full | Categorical representation learning and RG flow operators for algorithmic classifiers |
| title_fullStr | Categorical representation learning and RG flow operators for algorithmic classifiers |
| title_full_unstemmed | Categorical representation learning and RG flow operators for algorithmic classifiers |
| title_short | Categorical representation learning and RG flow operators for algorithmic classifiers |
| title_sort | categorical representation learning and rg flow operators for algorithmic classifiers |
| topic | renormalization group flow neural ODE hyperbolic geometry holographic duality category theory categorical representation learning |
| url | https://doi.org/10.1088/2632-2153/acb488 |
| work_keys_str_mv | AT artansheshmani categoricalrepresentationlearningandrgflowoperatorsforalgorithmicclassifiers AT yizhuangyou categoricalrepresentationlearningandrgflowoperatorsforalgorithmicclassifiers AT wenbofu categoricalrepresentationlearningandrgflowoperatorsforalgorithmicclassifiers AT ahmadrezaazizi categoricalrepresentationlearningandrgflowoperatorsforalgorithmicclassifiers |