简介：

简介：Thispaperproposesanactionanalysisforimplementingcombingpartialevaluationefficiently.Byanalyzingtheresultsofbindingtimeanalysis.Operations,whichshouldbeusedinthecombiningpartialevaluation.aredeterminedinadvance,sothatthecomputationinthecombinationofspecializedprogramsisreducedeffectively.

简介：

简介：

简介：亚群发现是关于利息的一个性质在不同变量之中发现有趣的协会的一种数据采矿技术。存在亚群发现方法为寻找采用不同策略，修剪并且评价亚群。学习哪个一个亚群发现算法的特征应该为产生优秀亚群被考虑是很关键的。在这方面，很多评论在亚群发现上被进行了。尽管他们在一些流行亚群发现方法上提供宽广概述，他们为亚群评估采用很少数据集和措施。根据存在措施，亚群不能从所有观点被估价。我们的工作由使用大量数据集并且由为亚群评估定义新措施在一些流行亚群发现方法上执行广泛的分析。分析结果将帮助做好理解主要亚群发现方法，揭开为进一步的改进的差距并且为特定的申请领域选择算法的合适的范畴。

简介：Parameterizationoftrianglemeshesisafundamentalproblemfortexturemapping,surfacefitting,surfacereconstruction,andmeshediting.Thedeformationoftriangularmeshescausedbytheparameterizedprocessisthemeasurementofparameterization.Traditionalstandardmethodhasitslimitationwhenevaluatingmixturedistortionenergyparameterizations.Thusanevaluationmethodbasesondistortionenergyparameterizationoftriangularmeshesisintroducedforthelimitation.Thenovelmethodemploysanadaptiveexpressionformtothemixtureenergy,andusesaweightfactortorepresentdistortionenergydistribution.Byusingthismethod,wecanevaluateallkindsofparameterizationinauniformmeasurementandacquireamoreintuitiveandclearevaluation.

简介：

简介：

简介：

简介：采用回归SVM框架，这份报纸建议语言学激发的特征工程策略与与人的评价的更好的关联开发一个MT评估度量标准。与所有可得到的特征的贪婪联合的当前的惯例相对照，六个特征为翻译质量根据人的直觉被建议。然后，语言学特征的贡献经由爬山的策略被检验并且分析。实验显示与彻底的语言学特征上的评价SVM模型或以前的尝试相比，有六个语言学信息的回归SVM模型基于特征更好概括到对面不同的数据集，并且与合适的非语言学的度量标准扩充这些语言学特征能完成另外的改进。

简介：

简介：Modelofellipticaltrainerand3DhumanbodyarebuiltwiththeergonomicmoduleofCATIAV5tostudyergonomiccharacteristicsoffitnessequipmentsasapreliminaryevaluationofellipticaltrainer.

简介：Thispapergivesamethodofquantifyingsmallvisualdifferencesbetween3Dmeshmodelswithconformingtopology,basedonthetheoryofstrainfields.Strainfieldisageometricquantityinelasticitywhichisusedtodescribethedeformationofelastomer.Inthispaperweconsiderthe3Dmodelsasobjectswithelasticity.Thefurtherdemonstrationsareprovided:thefirstisintendedtogivethereaderavisualimpressionofhowourmeasureworksinpractice;andthesecondistogivereadersavisualimpressionofhowourmeasureworksinevaluatingfilteralgorithms.Ourexperimentsshowthatourdifferenceestimatesarewellcorrelatedwithhumanperceptionofdifferences.Thisworkhasapplicationsintheevaluationof3Dmeshwatermarking,3Dmeshcompressionreconstruction,and3Dmeshfiltering.

简介：

简介：Withseveralricegenomeprojectsapproachingcompletiongeneprediction/findingbycomputeralgorithmshasbecomeanurgenttask.Twotestsetswereconstructedbymappingthenewlypublished28,469full-lengthKOMEricecDNAtotheRGPBACclonesequencesofOryzasativassp.japonica:asingle-genesetof550sequencesandamulti-genesetof62sequenceswith271genes.Thesedatasetswereusedtoevaluatefiveabinitiogenepredictionprograms:RiceHMM,GlimmerR,GeneMark,FGENSHandBGF.Thepredictionswerecomparedonnucleotide,exonandwholegenestructurelevelsusingcommonlyacceptedmeasuresandseveralnewmeasures.Thetestresultsshowaprogressinperformanceinchronologicalorder.Atthesametimecomplementarityoftheprogramshintsonthepossibilityoffurtherimprovementandonthefeasibilityofreachingbetterperformancebycombiningseveralgene-finders.

简介：Currently,daylightingductssystemiswidelyusedasadaylightingdevice.Generally,daylightingductsystemefficientlytakeslightfromoutsideduringtheday,andconveysdaylighttorequiredlocationthroughlightductmanufacturedbyhighreflectancemirror.Daylightingductsystemcanconveydaylighttoundergroundspacethathasnowindowsopeningtoexternalspace.Daylightingsystemiscomposedoflightcollectionpart,lightguidepartandlightemissionpart.Efficiencyofdaylightingsystemisdependingontypeofeachpartusedinthesystem.However,itisverydifficulttoestimateexactlightflowinthesystemconsideringtypeoftheparts.Authorsperformedmeasurementexperimentstomakeclearthelightflowwithreal-sizemodelandminiaturemodelofdaylightingductsystem.Wediscussedeffectoftypeofthepartsonefficiencyofdaylightingductsystem.

简介：

简介：

简介：

简介：