GRASP for continuous optimization
Maurício G. C. Resende, Celso Carneiro Ribeiro · 2016
Continuous GRASP, or C-GRASP, extends GRASP to the domain of continuous box-constrained global optimization. The algorithm searches the solution space over a dynamic grid. Each iteration of C-GRASP consists of two phases. In the construction (or diversification) phase, a greedy randomized solution is constructed. In the local search (or intensification) phase, a local search algorithm starts from the first phase solution and produces an approximate locally optimal solution. A deterministic rule triggers a restart after each C-GRASP iteration. This chapter addresses the construction phase and the restart strategy, and presents a local search procedure for continuous GRASP. These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.