A Graph-Neural-Network-Powered Solver Framework for Graph Optimization Problems
A Framework for Solving Graph Optimization Problems Based on Graph Neural Networks Background and Research Motivation In solving Constraint Satisfaction Problems (CSPs) and Combinatorial Optimization Problems (COPs), a common method is the combination of backtracking and branch heuristics. Although branch heuristics designed for specific problems a...