文档介绍：该【DNA微阵列数据的变量选择方法研究的中期报告 】是由【niuww】上传分享，文档一共【3】页，该文档可以免费在线阅读，需要了解更多关于【DNA微阵列数据的变量选择方法研究的中期报告 】的内容，可以使用淘豆网的站内搜索功能，选择自己适合的文档，以下文字是截取该文章内的部分文字，如需要获得完整电子版，请下载此文档到您的设备，方便您编辑和打印。,-test,foldchange,Lasso,,whichconsistsof54,,andgeneswithlowvariancewerefilteredout,leaving19,,witharatioof2:-uracy,precision,,,,,thehighdimensionalityofDNAmicroarraydatapresentsachallengefordataanalysis,-dimensionaldata,,includingthet-test,foldchange,Lasso,,(GEO)database(accessionnumberGSE39582).,whichconsistsof54,675probesetsrepresentingover47,(RMA)algorithm,whichincludesbackgroundcorrection,normalization,,leaving19,:t-test,foldchange,Lasso,-testandfold--testmeasuresthedifferenceinmeanexpressionlevelsbetweentwogroups,-(AUC)toevaluatethemodel',uracy,precision,-testandfoldchangemethodsintermsofAUC,accuracy,precision,andrecall(Table1).TheLassomethodselected18genes,,|Method|Numberofselectedgenes|AUC|Accuracy|Precision|Recall||--------|------------------------|-----|----------|-----------|--------||T-test|237||||||Foldchange|278||||||Lasso|18||||||RandomForests|14|||||ConclusionandFutureWorkInthisstudy,-uracy,precision,,theLassoandRandomForestsmethodsselectedlargelynon-overlappingsetsofgenes,,wewillexploreothervariableselectionmethodsandfeatureengineeringtechniquestofurtherimprovetheperformanceofthemodel.