组中选优机器学习问题建模和算法研究.pdf

Abstract Abstract In andsocial area numberof in fields，there nature，economy large samples taskforthis istolearntheruleof thebest classificationproblem groups．The selecting in onesfrom usetheruleto thebestonestheunseen groupdata(SBG)，then predict the forthe machine as data．While existing learningproblem group samples appear and machine considerdata the individual，fewlearningproblem group selecting froma data． optimalsample group Inviewofthe aboveintheactual work，this existingproblems paperputs anew of thebestonefrom data forwardmachine learningproblemselecting group firststudiesthe and forSBGmachine (SBG)．Thepaper modelingalgorithm learning forthenewmodelsand carried is out． problem．Then，aspecificapplication algorithm Theresultsforthe showtheeffectivenessand ofthenew application superiority as models．Themainworksofthis are follows： paper 1．SBGmachine anditsnewcharacteristicsare learningproblem pres