eaac语音编码算法的研究与实现.ppt

eaac语音编码算法的研究与实现.ppt

ID:55795600

大小:2.53 MB

页数:39页

时间:2020-06-07

eaac语音编码算法的研究与实现.ppt_第1页
eaac语音编码算法的研究与实现.ppt_第2页
eaac语音编码算法的研究与实现.ppt_第3页
eaac语音编码算法的研究与实现.ppt_第4页
eaac语音编码算法的研究与实现.ppt_第5页
资源描述:

《eaac语音编码算法的研究与实现.ppt》由会员上传分享,免费在线阅读,更多相关内容在应用文档-天天文库

1、WeirenYu1,JiajinLe2,XueminLin1,WenjieZhang1OntheEfficiencyofEstimatingPenetratingRankonLargeGraphs1UniversityofNewSouthWales&NICTA,Australia2DonghuaUniversity,ChinaSSDBM20122.ProblemDefinitionContents4.ExperimentalResults1.Introduction3.OptimizationTechniquesP

2、-Rank:ANewLink-basedSimilarityMeasureStructuralSimilarityMeasurePageRank[Pageet.al,1999]SimRank[JehandWidom,KDD02]P(enetrating)-RanksimilarityAnewpromisingstructuralmeasure[Zhaoet.al.,CIKM09]AnextensionofSimRankmetricsBasicPhilosophyTwoentitiesaresimilar,if (1

3、)theyarereferencedbysimilarentities (2)theyreferencesimilarentitiesP-RankOverviewFeaturesAvoiding“limitedinformationproblem”ofSimRank---Bytakingaccountofbothin-andout-linksDefinedrecursivelyandiscomputediterativelyApplicabletoanydomainwithobject-to-objectrelat

4、ionshipsChallengesCostlytocomputeP-RankonlargegraphsNaïveIterationO(Kn4)[Zhaoet.al.,CIKM09]PartialSumsAmortizationO(Kn3)[Lizorkinet.al.,PVLDB08]HardtoestimatetheerrorforP-RankapproximationRadius-andcategory-basedPruningRuleO(Kd2n2)[Zhaoet.al.,CIKM09]P-RankForm

5、ulationMathematicalFormulaIterativeParadigmContributionsCharacterizingP-Rankastwoformsmatrixinversion---deterministicoptimizationpowerseries---probabilisticcomputationDeterministicoptimization(off-line)eliminatingneighborhoodstructureredundancyquadratic-timewi

6、thanerrorboundProbabilisticcomputation(on-line)asamplingapproachlinear-timewithcontrolledaccuracyWhatisP-Rank?Thesimilarityinadomaincanbemodeledasgraphs. [verticesobjects,edgesrelationships]SimRankisanimportantsimilaritymeasurewhichexploitstherelationshipsbe

7、tweenverticesonwebgraphs.(GlenJeh&JenniferWidom,’02)Basicintuition:Twoobjectsaresimilariftheirneighborsaresimilar. (therecursivedefinition)Objectsaremaximallysimilartothemselves. (thebasecase)ExistingSimilaritymeasuresTextual-ContentSimilarity(text-based)Vecto

8、r-cosinesimilarity,PearsoncorrelationinIRStructural-ContextSimilarity(link-based)PageRankOnepage’sauthorityisdecidedbyitsneighbors’authorities.SimRankTwoobjectsaresimilariftheyarer

当前文档最多预览五页,下载文档查看全文

此文档下载收益归作者所有

当前文档最多预览五页,下载文档查看全文
温馨提示:
1. 部分包含数学公式或PPT动画的文件,查看预览时可能会显示错乱或异常,文件下载后无此问题,请放心下载。
2. 本文档由用户上传,版权归属用户,天天文库负责整理代发布。如果您对本文档版权有争议请及时联系客服。
3. 下载前请仔细阅读文档内容,确认文档内容符合您的需求后进行下载,若出现内容与标题不符可向本站投诉处理。
4. 下载文档时可能由于网络波动等原因无法下载或下载错误,付费完成后未能成功下载的用户请联系客服处理。