摘要
Thispaperproposesasupervisedtraining-testmethodwithGeneticProgramming(GP)forpatternclassification.Comparedandcontrastedwithtraditionalmethodswithregardtodeterministicpatternclassifiers,thismethodistrueforbothlinearseparableproblemsandlinearnon-separableproblems.Forspecifictrainingsamples,itcanformulatetheexpressionofdiscriminatefunctionwellwithoutanypriorknowledge.Atlast,anexperimentisconducted,andtheresultrevealsthatthissystemiseffectiveandpractical.
出版日期
2001年02月12日(中国期刊网平台首次上网日期,不代表论文的发表时间)