Simulated annealing with thresheld convergence

Stephen Chen*, Carlos Xudiera, James Montgomery

*Corresponding author for this work

    Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

    12 Citations (Scopus)

    Abstract

    Stochastic search techniques for multi-modal search spaces require the ability to balance exploration with exploitation. Exploration is required to find the best region, and exploitation is required to find the best solution (i.e. the local optimum) within this region. Compared to hill climbing which is purely exploitative, simulated annealing probabilistically allows "backward" steps which facilitate exploration. However, the balance between exploration and exploitation in simulated annealing is biased towards exploitation - improving moves are always accepted, so local (greedy) search steps can occur at even the earliest stages of the search process. The purpose of "thresheld convergence" is to have these early-stage local search steps "held" back by a threshold function. It is hypothesized that early local search steps can interfere with the effectiveness of a search technique's (concurrent) mechanisms for global search. Experiments show that the addition of thresheld convergence to simulated annealing can lead to significant performance improvements in multi-modal search spaces.

    Original languageEnglish
    Title of host publication2012 IEEE Congress on Evolutionary Computation, CEC 2012
    DOIs
    Publication statusPublished - 2012
    Event2012 IEEE Congress on Evolutionary Computation, CEC 2012 - Brisbane, QLD, Australia
    Duration: 10 Jun 201215 Jun 2012

    Publication series

    Name2012 IEEE Congress on Evolutionary Computation, CEC 2012

    Conference

    Conference2012 IEEE Congress on Evolutionary Computation, CEC 2012
    Country/TerritoryAustralia
    CityBrisbane, QLD
    Period10/06/1215/06/12

    Fingerprint

    Dive into the research topics of 'Simulated annealing with thresheld convergence'. Together they form a unique fingerprint.

    Cite this