Dragan Stevanović: Reinforcement learning on graphs: cross-entropy methods and basic graph environments

Adam Zsolt Wagner [arXiv:2104.14516] recently showed how reinforcement learning (RL) can be applied to construct (counter)examples in graph theory. We will showcase here a more readable, more stable and significantly faster reimplementation of his approach, and illustrate its work by finding counterexamples to a few published conjectures. We will also shortly discuss ways to implement several new RL environments that will cover constructions of simple graphs and trees, their signed variants, and graphs with bounded maximum vertex degree.

Join Zoom Meeting HERE!