跳到主要內容

臺灣博碩士論文加值系統

(44.211.34.178) 您好!臺灣時間:2024/11/12 14:44
字體大小: 字級放大   字級縮小   預設字形  
回查詢結果 :::

詳目顯示

: 
twitterline
研究生:雷德仁
研究生(外文):tejen lei
論文名稱:國防預算預測之研究:應用類神經網路與支援向量機
論文名稱(外文):The Study of the Forecasting of National Defense Budget:An Application of Artificial Neural Network and support vector machines
指導教授:王維康王維康引用關係左杰官左杰官引用關係
指導教授(外文):weikang wangBrandt Tso
學位類別:碩士
校院名稱:國防管理學院
系所名稱:國防財務資源研究所
學門:商業及管理學門
學類:財務金融學類
論文種類:學術論文
論文出版年:2006
畢業學年度:94
語文別:中文
論文頁數:62
中文關鍵詞:國防預算類神經網路支援向量機
外文關鍵詞:National defense budgetArtificial neural networkssupport vector machines
相關次數:
  • 被引用被引用:13
  • 點閱點閱:667
  • 評分評分:
  • 下載下載:225
  • 收藏至我的研究室書目清單書目收藏:1
近年來台灣的國防預算面臨成長趨緩、資源受限情事,為避免影響建軍備戰及國防安全。發展一套準確預測國防預算額度的模式,以先期完成適當配置實屬重要。而影響國防預算額度的因素,一般考量幾個層面,首先針對外部環境威脅,衡量實際戰備需求;接著顧及國內經濟現況、財務負荷、政治干預與預算制度的影響。國防預算預測屬時間序列的預測,目前已有多篇研究結果顯示類神經網路與支援向量機是財務時間序列模式建立與預測的有用工具。
故本研究將影響台灣國防預算預測之內、外部變數,應用類神經網路中的倒傳遞類神經網路與支援向量機,其中參數部分,類神經網路透過遺傳演算法、支援向量機使用五組交叉驗證格子法分別求取最適參數,資料來源為中華民國統計資訊網等,評量標準以最低的MSE、NMSE、MAE值及最大r值的模式,為最佳之國防預算預測模式。
實驗結果顯示,無論單以影響台灣國防預算的內部因素、外部因素或結合內外部因素,經過倒傳遞類神經網路等方法預測結果均可適當預測出國防預算趨勢,其中以影響國防預算內部因素為輸入變數,配合遺傳演算法結合倒傳遞類神經網路進行國防預算預測結果最為精確。期望藉由預算預測精確性的提升,增進國防預算編列的效能。
Taiwan’s national defense budget is easing up and leading resource allocation to be limited in recent years. It is very important to develop one model which predicts the national defense budget amount accurately. So it can allocate properly in advance and meet the safe demand of the national defense and the force preparing for war. Conventionally, people think that a countary’s defense budget depends on the threat of a rival country, its economic growth, its party affiliation and the system of the budget. The national defense budget prediction is regarded as a time-series prediction. A large number of successful applications have showed that artificial neural networks and support vector machines can very useful for time-series modeling and forecasting .
This study used the internal and external factors of national defense budget as input variables , back-propogation network and support vector machines to predict national defense budget. The best parameters of BPN and SVM are selected by Genetic Algorithms and 5-fold grid methods. The data set came from National Statistics, etc.. The best model of national defense budget is the lowest MSE,NMSE,MAE values and the highest r values.
The experimental result shows that internal or external or both factors of national defense budget could predict the trend of national defense properly with the tools such as neural networks, etc.. The best prediction performance of GA-BPN is produced when the input variables are the internal factors of national defense budget. The improvement of the accuracy with the budget is expected to promote the efficiency of national defence budget.
第ㄧ章 緒論.................................................1
1.1 研究背景與動機.......................................1
1.2 研究目的............................................3
1.3 論文架構............................................4
第二章 文獻探討.............................................5
2.1 國防預算理論........................................5
2.2 國防預算影響因素....................................6
2.3 國防預算預測方法...................................13
2.4 類神經網路於商業預測的應用.........................15
2.5 支援向量機於商業預測的應用.........................20
第三章 研究方法............................................22
3.1 遺傳演算法.........................................22
3.2 類神經網路.........................................26
3.3 支援向量機.........................................33
第四章 實證結果分析........................................40
4.1 倒傳遞類神經網路預測國防預算.......................42
4.2 遺傳演算法結合倒傳遞類神經網路預測國防預算.........51
4.3 支援向量機預測國防預算.............................54
4.4 國防預算預測結果比較...............................58
第五章 結論與建議..........................................61
5.1 研究結論...........................................61
5.2 後續研究建議.......................................62
參考文獻....................................................63
[1] 楊志清(2000),「時間數列模型在國防預算編列上之應用」,淡江大學統計學系碩士論文。
[2] 楊開銘(2000),「中國國防費估測模式之研究」,國防管理學院資源所碩士論文。
[3] 陳仁龍(2001),「軍事投資最適化比例之研究」,國防管理學院決策所碩士論文。
[4] 劉裕國(2001),「後冷戰時期影響我國防預算系統之國內外政經因素分析」,成功大學政治經濟研究所碩士論文。
[5] 陳建達(2002),「中國經濟成長、國防費與軍事能力關係之研究」, 國防管理學院資源所碩士論文。
[6] 陳章仁(2002),「威脅型態、軍事防禦能力、區域衝突與國防預算關係之研究」,國防管理學院財管所碩士論文。
[7] 吳智鴻(2003),「使用遺傳演算法最佳化支援向量機參數-財務危機預警之應用」,台北大學商管所博士論文。
[8] 國防部(2004),「中華民國九十三年國防報告書」,網址http://report.mnd.gov.tw/。
[9] 王界人、陳穆臻、黃承龍(2004),「支援向量機於信用評等之應用」,計量管理期刊,第1卷,第2期, 155-172頁。
[10]葉怡成(2004),「類神經網路模式應用與實作」,儒林書局。
[11]劉易昌(2004),「支援向量機於財務預測上之應用」,靜宜大學資訊管理學系碩士論文。

[1] Coetzee,C. [2002],”SOUTH AFRICAN DEFENCE EXPENDITURE IN THE 20TH CENTURY”University of Stellenbosch.
[2] Cortes,C.and Vapnik,V.,[1995],”Support vector networks”Machine Learning,No.20,pp.273-297
[3] Duan,K.S.et al.[2003],”Evaluation of simple performance measures for tuning SVM hyperparameters” Neurocomputing,No.51,pp.41-59.
[4] Gunn,S.R.[1998], ”Support Vector machines for classification and regression,”Technical Report, University of Southampton.
[5] Hamid,S.A.and Habib,A. [2005],“CAN NEURAL NETWORKS LEARN THE BLACK-SCHOLES MODEL?: A SIMPLIFIED APPROACH” Working Paper No. 2005-01 School Of Business, Southern New Hampshire University, 2500 N. River Road, Manchester, NH 03106
[6] Hsu,C.W.et al.[2003],”A Practical Guide to Support Vector Classification” Technical Report,Department of Comptuer Science & Information Engineering, National Taiwan University.
[7] Kim,H.C.[2001],”Determinants of U.S. Military Expenditure After the Cold War” University of Missouri - Columbia, Dept. of Pol.Sci.
[8] Kwon,Y.[2005],”Nonlinear Feature Extraction Using a Neuro Genetic Hybrid” GECCO 2005, pp.2089-2096,Washington, DC, USA.
[9] Kwon,Y.et al.[2005],”Stock Prediction Based on Financial Correlation” GECCO 2005, pp.2061-2064,Washington, DC, USA.
[10]Lin,C.J.[2005],”LIBSVM -- A Library for Support Vector Machines”,http://www.csie.ntu.edu.tw/%7Ecjlin/libsvm/index.html
[11]Markey,M.K.et al.[2005],“Impact of missing data in evaluating artificial neural networks trained on complete data” Computers in Biology & Medicine,Vol.35,pp.1-10.
[12]Richardson,L.F.[1960],”Arms and Insecurity: A Mathematical Study of the Causes and Origins of War”Pittsburgh: Boxwood Press.
[13]Skabar,A.and Cloete,I. [2002],”Neural Networks,Financial Trading and the Efficient Markets Hypothesis” Proceedings of the twenty-fifth Australasian conference on Computer science ,Vol.4,pp.241-249.
[14]Smola,A.and Vapnik,V. [1996]“Support Vector Regression Machines”Bell Labs and Monmouth University Department of Electronic Engineering West Long Branch, NJ 07764, USA
[15]Tay,E.H.and Cao,L. [2001],”Application of support vector machines in financial time series forecasting” ,Omega,Vol.29,pp.309-317.
[16]Yang,H.et al.[2003],” Financial Time Series Prediction Using Non-fixed and Asymmetrical Margin Setting with Momentum in Support Vector Regression”International Conference on Neural Information Processing
[17]Yu,T.C.[2002],”THE IMPACT OF US-CHINA RELATIONS ON TAIWAN’S MILITARY SPENDING: AN ANALYTICAL MODEL OF ERROR CORRECTION MECHANISM (1966-1992)”CGOTS,APSA 2002 annual meeting,Boston.
QRCODE
 
 
 
 
 
                                                                                                                                                                                                                                                                                                                                                                                                               
第一頁 上一頁 下一頁 最後一頁 top