资源描述:
《Copulas and time series with long-ranged dependencesCopulas与长程相关时间序列》由会员上传分享,免费在线阅读,更多相关内容在学术论文-天天文库。
1、Copulasandtimeserieswithlong-rangeddependencesRemyChicheporticheandAnirbanChakrabortiyChairedenancequantitative,EcoleCentraleParis,92295Ch^atenay-Malabry,France(Dated:July16,2018)Wereviewideasontemporaldependencesandrecurrencesindiscretetimeseriesfromseveralareasofnaturalandsoci
2、alsciences.Werevisitexistingstudiesandredenetherelevantobserv-ablesinthelanguageofcopulas(jointlawsoftheranks).Weproposethatcopulasprovideanappropriatemathematicalframeworktostudynon-lineartimedependencesandrelatedconcepts
3、likeaftershocks,Omorilaw,recurrences,waitingtimes.Wealsocri
4、ticallyargueusingthisglobalapproachthatpreviousphenomenologicalattemptsinvolvingonlyalong-rangedautocorrelationfunctionlackedcomplexityinthattheywereessentiallymono-scale.PACSnumbers:05.45.Tp,02.50.-r,87.85.NgKeywords:Recurrenceintervals,copulas,long-rangedcorrelations,timeseriesI.I
5、NTRODUCTIONetal.[5]hadsuggestedauniedscalinglawcombiningtheGutenberg-Richterlaw,theOmorilaw,andthefrac-Athoroughunderstandingoftheoccurencesandtaldistributionlawinasingleframework.Thisglobalstatisticsofextremeeventsiscrucialineldslikeseis-approachwaslaterextendedbyCorral[18,19],wh
6、opro-micity,nance,astronomy,physiology,etc.[1,2].TheposedtheexistenceofauniversalscalinglawforthePDFanalysesofextremeeventsplaysapivotalroleeverytimerecurrencetimesbetweenearthquakesinagivenregion.anaddressedproblemhasastochasticnature,sincetheThisisusefulbecause,duetothescalingpro
7、perties,itrareextremeeventscanhaveratherstrongordrasticispossibletoanalysethestatisticsofreturnintervalsconsequences
8、makingitwidelyuseful.Onetheoreti-fordierentthresholdsbystudyingonlythebehaviorofcalmotivationforstudyingextremeeventsinaparticu-small
uctuationsoccurringveryfrequent
9、ly,whichhavelareldlikenance,istoaccountfortheobservedfatmuchbetterstatisticsandreliabilitythanthoseofthetailsoflog-returns(deviationfromtheNormaldistri-rareextremelarge
ucutations.Italsorevealsaspa-butioninthetails)ofstockprices[3].Amorepracti-tiotemporalorganizationoftheseismicit
10、y,assuggestedcalmot