6533b7d0fe1ef96bd125ab76
RESEARCH PRODUCT
Intelligent Multi-Start Methods
Abraham DuarteRicardo AcevesJose M. Moreno-vegaMaria Teresa LeónRafael Martísubject
Mathematical optimization021103 operations researchOptimization problemDegree (graph theory)Computer sciencemedia_common.quotation_subject0211 other engineering and technologiesCombinatorial optimization problem020206 networking & telecommunications02 engineering and technologyDiversification (marketing strategy)0202 electrical engineering electronic engineering information engineeringQuality (business)Metaheuristicmedia_commondescription
Heuristic search procedures aimed at finding globally optimal solutions to hard combinatorial optimization problems usually require some type of diversification to overcome local optimality. One way to achieve diversification is to re-start the procedure from a new solution once a region has been explored, which constitutes a multi-start procedure. In this chapter we describe the best known multi-start methods for solving optimization problems. We also describe their connections with other metaheuristic methodologies. We propose classifying these methods in terms of their use of randomization, memory and degree of rebuild. We also present a computational comparison of these methods on solving the Maximum Diversity Problem to illustrate the efficiency of the multi-start methodology in terms of solution quality and diversification power.
| year | journal | country | edition | language |
|---|---|---|---|---|
| 2018-09-21 |