Evolution of neural networks topologies and learning parameters to produce hyper-heuristics for constraint satisfaction problems
José Carlos Ortíz-Bayliss, Hugo Terashima‐Marín, Peter D. Ross, Santiago Enrique Conant-Pablos · 2011
This paper describes a model which constructs hyper-heuristics for variable ordering within Constraint Satisfaction Problems (CSPs) by running a genetic algorithm that evolves the topology of neural networks and some learning parameters.