Current browse context:
cs.LG
Change to browse by:
References & Citations
Computer Science > Machine Learning
Title: Learning Combinatorial Optimization on Graphs: A Survey with Applications to Networking
(Submitted on 22 May 2020 (v1), last revised 13 Jul 2020 (this version, v2))
Abstract: Existing approaches to solving combinatorial optimization problems on graphs suffer from the need to engineer each problem algorithmically, with practical problems recurring in many instances. The practical side of theoretical computer science, such as computational complexity, then needs to be addressed. Relevant developments in machine learning research on graphs are surveyed for this purpose. We organize and compare the structures involved with learning to solve combinatorial optimization problems, with a special eye on the telecommunications domain and its continuous development of live and research networks.
Submission history
From: Natalia Vesselinova [view email][v1] Fri, 22 May 2020 09:45:36 GMT (3600kb,D)
[v2] Mon, 13 Jul 2020 18:03:39 GMT (3727kb,D)
Link back to: arXiv, form interface, contact.