LGCVMLMay 15, 2019

A Learning based Branch and Bound for Maximum Common Subgraph Problems

arXiv:1905.05840v24 citations
Originality Incremental advance
AI Analysis

This work addresses a specific bottleneck in combinatorial optimization for researchers and practitioners, but it is incremental as it builds on existing branch-and-bound methods.

The paper tackled the problem of improving branch-and-bound algorithms for maximum common subgraph problems by developing a reinforcement learning-based branching heuristic to reduce search tree size, and experiments showed it outperforms the current best algorithm.

Branch-and-bound (BnB) algorithms are widely used to solve combinatorial problems, and the performance crucially depends on its branching heuristic.In this work, we consider a typical problem of maximum common subgraph (MCS), and propose a branching heuristic inspired from reinforcement learning with a goal of reaching a tree leaf as early as possible to greatly reduce the search tree size.Extensive experiments show that our method is beneficial and outperforms current best BnB algorithm for the MCS.

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes