Add to Chrome
Add to Firefox
Add to EdgeGradient Descent based Weight Learning for Grouping Problems: Application on Graph Coloring and Equitable Graph Coloring
Sep 05, 2019
A grouping problem involves partitioning a set of items into mutually disjoint groups or clusters according to some guiding decision criteria and imperative constraints. Grouping problems have many relevant applications and are computationally difficult. In this work, we present a general weight learning based optimization framework for solving grouping problems. The central idea of our approach is to formulate the task of seeking a solution as a real-valued weight matrix learning problem that is solved by first order gradient descent. A practical implementation of this framework is proposed with tensor calculus in order to benefit from parallel computing on GPU devices. To show its potential for tackling difficult problems, we apply the approach to two typical and well-known grouping problems (graph coloring and equitable graph coloring). We present large computational experiments and comparisons on popular benchmarks and report improved best-known results (new upper bounds) for several large graphs.
