简介:OutlierminingisanimportantaspectindataminingandtheoutlierminingbasedonCookdistanceismostcommonlyused.Butweknowthatwhenthedatahavemulticollinearity,thetraditionalCookmethodisnolongereffective.Consideringtheexcellenceoftheprincipalcomponentestimation,weuseittosubstitutetheleastsquaresestimation,andthengivetheCookdistancemeasurementbasedonprincipalcomponentestimation,whichcanbeusedinoutliermining.Atthesametime,wehavedonesomeresearchonrelatedtheoriesandapplicationproblems.