We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

math.CO

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Mathematics > Combinatorics

Title: Constructions in combinatorics via neural networks

Abstract: We demonstrate how by using a reinforcement learning algorithm, the deep cross-entropy method, one can find explicit constructions and counterexamples to several open conjectures in extremal combinatorics and graph theory. Amongst the conjectures we refute are a question of Brualdi and Cao about maximizing permanents of pattern avoiding matrices, and several problems related to the adjacency and distance eigenvalues of graphs.
Comments: 23 pages, 13 figures
Subjects: Combinatorics (math.CO); Machine Learning (cs.LG)
Cite as: arXiv:2104.14516 [math.CO]
  (or arXiv:2104.14516v1 [math.CO] for this version)

Submission history

From: Adam Zsolt Wagner [view email]
[v1] Thu, 29 Apr 2021 17:32:56 GMT (1415kb,D)

Link back to: arXiv, form interface, contact.