模糊粗糙多层风险评估算法与应用.pdf

英文摘要 ABSTRACT Unreliableconclusionsareoften drawninriskassessmenttasksduetothe existenceofdifferentof inthe typesuncertaintydata．Byintroducingfuzzy．rough attribute toextreme weights fl learningmachine(ELM)method，thispresents paper inalleffort fuzzy·rough to the weightedELM(FRW—ELM)algorithm exploit informationembeddedinthedata． uncertainty Thismethod a hierarchicalrisk proposesfuzzy-rough assessmentthat approach FRW-ELMandthe integrates recently developedfuzzyboundaryregion—based feature tohandle data fuzzy—roughselection(B—FRFS)techniquehigh—dimensional thisrisk assessment ofeach nodeinthe effectively．In approach，thesignificanceinput ELMisinduced its contmnedwithinthe ofthe by uncertaintydegree boundaryregion associated measuresofallthe fuzzy—roughset．Importantly，therequireduncertainty attributesareobtainedasthe ofthefeatureselectionthat by·product process implements B-FRFS．InSOdoing，the hierarchicalriskassessm