polynomial time

approximation algorithms

traveling salesman

spanning tree

assignment problem

perfect matchings

graphs

graph theory

computational experiments

graph coloring

IEEE Access

Existing approaches to solving combinatorial optimization problems on graphs\nsuffer from the need to engineer each problem algorithmically, with practical\nproblems recurring in many instances. The practical side of theoretical\ncomputer science, such as computational complexity, then needs to be addressed.\nRelevant developments in machine learning research on graphs are surveyed for\nthis purpose. We organize and compare the structures involved with learning to\nsolve combinatorial optimization problems, with a special eye on the\ntelecommunications domain and its continuous development of live and research\nnetworks.\n

Learning Combinatorial Optimization on Graphs: A Survey With Applications to Networking