長久以來,人類在萬物中獨具的高智慧特質吸引了無數的哲學家和科學家

投入對其研究,除了醫學的原因之外,由於人腦所具有卓越的辨識系統及學

習能力,為數不少的科學家們相信人腦存在許多最適化系統與設計,因此如

何模仿人類腦神經的組織與運作,一直是很多人努力及夢寐以求的.因此類

神經網路就是依據這些理念而在各研究領域上廣為發展與應用,其中本文

所探討的倒傳遞神經網路模型更是目前類神經網路模型中最具代表性,應

用最廣的模型.而結構性時間數列模型則是將可被觀察的變數分解成趨勢,

季節性,不規則性等不可被觀察項,故其對經濟意義的解釋是相當明當明顯

的,藉由狀態空間模式的轉換,我們將很容易地利用卡門濾器來作估計與預

測.而本文所欲探的重點在於比較有學習機能的倒傳遞神經網及可利用最

新的資訊更新之結構性時間數列何者之預測能利較佳,藉此瞭解二者之一

些特性.

吳柏林、賴家瑞、劉勇衫，台灣地區外籍觀光旅客意願分析與預
測模式，民國八十一年
葉怡成，類神經網路模式應用與質作，儒林圖書，民國八十二年
元月
焦成李，神經網路系統必z 論， 儒林圖書，民國八十年十月

Diebold,F.X.(l989),"SLructural Time Series Analysis and iVfodelling :
Package A Review", Journal 0/ Applied Econometrics, vo1.4,PP195-
204.
Harvey,A.C.and S.Peters(1990),"Estimation Procedures for structural
Time series lvIoclels" ,Journal of Forecasling,vo\.9 ,PP89-108.
Harvey,A.C.(19S9), 'TorccasLing,structural Times series models and
the Kalman Filter" ,Cambridge University Press,Cambridge.
Harvey,A.C.(1981), "Times Series moclcls",oxforcl : Phillip Allan and
Atlantic Highlancls,NJ : Ilum(lntics Press.
Harvey,A.C.(1990), "The Econometric Anlysis of Time Series" ,Journal
of Forecasting,voI.9,PPS9-108.
Harvey,A.C.(l984), "A Unified View of ststistical forecasting Proce-
dures ",Journal of FOl'ccasting,3,PP245-2S3.
I-Iarvey,A.C.(19S5), "Trends and Cycles in Macroeconomic Times series"
,Journal of Bllssincss and Economic Slalistics,3,PP21G-227.
Harvey,A.C. and J.Durbin(1986), "The effects of scat Belt Legislation
on British Roael Casualties: A case study in structural series J\tfodelling"
,Journal of The Royal Statistical socicty}series A ,PP187-227.
I-Iarvey,A .C.,Hcnry B.,Pcters,S and S wren-Lewis(1986), "Stochastic
Trends in Dynamic Regression ~' [oclels : ;\ n Application to The empolymentoutput
equation" , Eco/l.omic JO -llrnal,9G,PP975-985.
Kitagawa,G and W Gersch( 198'1)," A Smoothness Priors-State Space
Modelling of Time series with Trend ancl seasonality", JouTnal of the
American Statistical !lssocialion,79,PP:378-389.
Freeman,J.A. and Skapura,D .M.(1991 )," Neural Networks-A Igorithms,Applications
,and Programming Techniques" ,Addison '-\fesley.
vVeigencl,A.S.,llumclhart,D.E. and Huberman,B.A.( -) "Back-Propagation
,vVeight-Elimination and Time Series Prediction".
Harvey,A.C. anel Toclel,P.(1983),'Torecasting economic Time Series with
Structural and Box-Jenkins Models: ;\ Case study" ,Journal of Bussiness
and Economic Slnlistics,PP299-:315.
Chatfield, C. ( 1987), "The lTol t- Wintel'S Porccasti ng Proceciure", Applied
Statistics, 2 7, P P2G•I-:279.
Brown,R.G.(1963),”Smoothing ,Forecasting and Prediction”, Englewood Cliffs: Prentice Hall.
Maravall,A.(1985),”On Structural Time Series Models and The Characterization of Components”, Journal of Business and Economic Statistics3(4), pp350-55.
William H.G.(1991),”Econometric Analysis”,Maxwell Machillan.
Makridakis,S.,Andersen,A.,Carbone,R.,Fildes,R.,Hibun,M.,Lewandowski,R.,Newton,J.,Parzon,E. and Winter,R.(1982), “The Accuracy of Extrapolation(time Series)Methods: Result of Forecasting Competition”, Journal of Forecasting ,PP111-153.
Judge,G.G.,Hill,R.C.,Griffiths, W.E.,Lutkepohl,H. and Lee ,T.C.(1988),”Introduction to The Theory and Practice of Econometrics”, Wie,Wiley.