pith. sign in

arxiv: 1905.11623 · v2 · pith:OVJ3SOYNnew · submitted 2019-05-28 · 💻 cs.LG · stat.ML

Solving NP-Hard Problems on Graphs with Extended AlphaGo Zero

classification 💻 cs.LG stat.ML
keywords problemsgraphlearningmethodalphagobettercombinatorialframework
0
0 comments X
read the original abstract

There have been increasing challenges to solve combinatorial optimization problems by machine learning. Khalil et al. proposed an end-to-end reinforcement learning framework, S2V-DQN, which automatically learns graph embeddings to construct solutions to a wide range of problems. To improve the generalization ability of their Q-learning method, we propose a novel learning strategy based on AlphaGo Zero which is a Go engine that achieved a superhuman level without the domain knowledge of the game. Our framework is redesigned for combinatorial problems, where the final reward might take any real number instead of a binary response, win/lose. In experiments conducted for five kinds of NP-hard problems including {\sc MinimumVertexCover} and {\sc MaxCut}, our method is shown to generalize better to various graphs than S2V-DQN. Furthermore, our method can be combined with recently-developed graph neural network (GNN) models such as the \emph{Graph Isomorphism Network}, resulting in even better performance. This experiment also gives an interesting insight into a suitable choice of GNN models for each task.

This paper has not been read by Pith yet.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Neural optimization for quantum architectures: graph embedding problems with Distance Encoder Networks

    quant-ph 2026-05 unverdicted novelty 5.0

    A modified autoencoder with a custom embedding loss learns spatial mappings to solve the constrained unit disk problem for qubit embedding on neutral-atom quantum processors and outperforms classical solvers under fix...