模糊时间序列预测

Available online at doc.guandang.net
MathematicsandComputersinSimulation79(2008)

539–554
Acomputationalmethodofforecastingbasedon
fuzzytimeseries
S.R.Singh
DepartmentofMathematics,BanarasHinduUniversity,Varanasi221005,India
Received6February2008;accepted24February2008
Availableonline7March2008
Abstract
Inthispaper,acomputationalmethodofforecastingbasedonfuzzytimeserieshavebeendevelopedtoprovideimprovedforecastingresultstocopeupthesituationcontaininghigheruncertaintyduetolarge uctuationsinconsecutiveyear’svaluesinthetimeseriesdataandhavingnovisualizationoftrendorperiodicity.Theproposedmodelisoforderthreeandusesatimevariantdifferenceparameteroncurrentstatetoforecastthenextstate.Thedevelopedmodelhasbeentestedonthehistoricalstudentenrollments,UniversityofAlabamatohavecomparisonwiththeexistingmethodsandhasbeenimplementedforforecastingofacropproductionsystemoflahicrop,containinghigheruncertainty.Thesuitabilityofthedevelopedmodelhasbeenexaminedincomparisonwiththeothermodelstoshowitssuperiority.
©2008IMACS.PublishedbyElsevierB.V.Allrightsreserved.
Keywords:Fuzzytimeseries;Timeinvariant;Timevariant;Linguisticvariables;Fuzzylogicalrelations
1.Introduction
Thetimeseriesforecastinginvestigatestherelationsonthesequentialsetofpastdatameasuredovertimetoforecastthefuturevalues.Theareahasbeenwidelystudiedandtraditionalforecastingarefrequentlyconductedbystatisticaltoolslikeregressionanalysis,movingaverages,integratedmovingaverageandautoregressiveintegratedmovingaverage.Oneofthemajorlimitationsofthesemethodsisnotattendingtheforecastingproblemsinwhichthehistoricaldataareinlinguisticterms.FuzzysettheoryandfuzzylogicintroducedbyZadeh[28,29]providesageneralmethodforhandlinguncertaintyandvaguenessininformationavailableinlinguisticterms.SongandChissom[18–20]usedthefuzzysettheorygivenbyZadehtodevelopmodelsforfuzzytimeseriesforecastingandconsideredtheproblemofforecastingenrollmentsonthetimeseriesdataofUniversityofAlabama.Chen[1]presentedamethodoffuzzytimeseriesforecastingofenrollmentsusingthesimpli edarithmeticoperations.Themajorprobleminfuzzytimeseriesforecastingistheproblemofaccuracyinforecast.Manyresearchers,SullivanandWoodall[22];KimandLee[9];Hwangetal.[8];ChenandHwang[3];Huarng[6]workedonvariousmodelsforfuzzytimeseriesforecastingtoimprovetheforecast.
Tel.:+915422307435.
E-mailaddress:singhshivaraj@rediffmail.com.
0378-4754/$32.00©2008IMACS.PublishedbyElsevierB.V.Allrightsreserved.doi:10.1016/j.matcom.2008.02.026
The present paper proposes a new method of fuzzy time series forecasting based on difference parameters. The proposed method is a simpli?ed computational ...
A generalized method for forecasting based on fuzzy time series_自然科学_...uence forecasting accuracy. To reduce the computational time, Chen (1996) ...
Section 4 presents a concrete method of ...cant overload burden and leading to computational ...forecasting based on fuzzy time series and ...
Forecasting time series ... 7页 免费 A computational method o... 16页...based on combination of two techniques: fuzzy transform and perception-based ...
method for forecasting enrollments based on fuzzy time series and particle swarm optimization I-Hong Kuo a, Shi-Jinn Horng a,b,c,f,*, Tzong-Wann Kao...
A bivariate fuzzy time series(2008Huarng)_电子/...A two factor method to forecast temperature and ...forecasting based on fuzzy logical relationships and...
a simple method to forecast the unknown patterns....The forecasting results of both models are then ...based on fuzzy time series, Fuzzy Sets and ...
Fuzzy Time Series Based on Particle Swarm ...a new method to construct fuzzy logic ...forecasting accuracy or reduce the computational ...
《Average- Based Fuzzy time series models for forecasting Shanghai Compound ...《A New Method to Forecast Enrollments Using Fuzzy Time Series》 ...
a new method for load forecasting—fuzzy rules ...a computational analogy of adaptive systems, GAs ...Using Artificial Neural Networks and Time Series. ...
热门文档