简介:Inthispaperwestudytheproximalpointalgorithm(PPA)basedpredictioncorrection(PC)methodsformonotonevariationalinequalities.Eachiterationofthesemethodsconsistsofapredictionandacorrection.ThepredictorsareproducedbyinexactPPAsteps.ThenewiteratesarethenupdatedbyacorrectionusingthePPAformula.Wepresenttwoprofitfunctionswhichservetwopurposes:Firstweshowthattheprofitfunctionsaretightlowerboundsoftheimprovementsobtainedineachiteration.Basedonthisconclusionweobtaintheconvergenceinexactnessrestrictionsforthepredictionstep.Secondweshowthattheprofitfunctionsarequadraticallydependentuponthesteplengths,thustheoptimalsteplengthsareobtainedinthecorrectionstep.Inthelastpartofthepaperwecomparethestrengthsofdifferentmethodsbasedontheirinexactnessrestrictions.