欢迎来到天天文库
浏览记录
ID:32137071
大小:1.77 MB
页数:43页
时间:2019-01-31
《基于自相关函数的模糊时间序列模型的优化算法》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、万方数据英文摘要AnOptimizationAlgorithmforFuzzyTimeSeriesModelBasedonAutocorrelationFunctionAbstractFuzzytimesseriesmodelisapowerfultooltodealwiththehistoricaldatawhicharelinguisticvMuesanduncertaincircumstances,moreandmoreresearchersfocusonfuzzytimeseriesmodel.Atpresent,fuzzytimeseriesmodelhasbeenappliedt
2、oforecastthenumberoftourists,temperature,stockmarket,networkusers,incidenceofhemorrhagicfeverwithrenalsyndromeetc.Inordertogethigherprecision,differentfuzzytimeseriesmethodsareproposed.Moststudiesfocusontheuniverseofdiscoursepartitionanddatafuzzificationinfuzzytimeseriesmodels.Fuzzytimeseriesisevol
3、vedfromtheclassictimeseries,inwhichdatacorrelationisinvolvedinevitably.Thecorrelationwouldbenotchangedthroughdatafuzzification,resultingintheantecedentoffuzzyrulesinfluencingontheconsequentoffuzzyrulesinadifferentway,Thenthepaperproposesanewoptimizationalgorithmtooptimizefuzzyinferencerulesbycombin
4、ingtheautocorrelationtheoryoftimeseries.Firstly,stationarityofdataisdiscussedbyvirtueoftimeseriestheory.Ameannon。stationarysequenceistransformedintoastationarysequencethroughappropriatedifference.Inaddition,appropriatevariancestabilizingtransformationisrequiredforvariancenon—stationarysequence.Gene
5、rally,forthemeanandvariancenon—stationaryprocess,thelogarithmictransformationisappliedfirstlyandthenthedifferencetransformationisapplied.Basedonabovestep,fuzzysetsareobtainedthroughtraditionalfuzzysetsontheuniverseofdiscourseU,therebyfuzzyrulesareconstructed;Secondly,fuzzyrulesareoptimizedbyautocor
6、relationtheory.Atthesametime,animprovedmethodofstandardizeddeviationisestablishedforweightcalculation;Finally,throughcombiningwiththeprevioustimeseriesmodelandtraditionalfuzzytimeseriesmodelabouttheforecastingofAlabamauniversityenrollments,thecomparativeresultsshowthattheproposedmethodiseffective.万
7、方数据Keywords:Fuzzytimeseries;Autocorrelationfunctions;RulesWeights;Characteristicexpansionmethod.万方数据目录第l章绪论⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯11.1引言⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯.11.2存在的问题与本文的主要工作⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯⋯
此文档下载收益归作者所有