學術產出-Theses
Article View/Open
Publication Export
-
題名 應用類神經網路方法於金融時間序列預測之研究--以TWSE台股指數為例
Using Neural Network approaches to predict financial time series research--The example of TWSE index prediction作者 張永承
Jhang, Yong-Cheng貢獻者 姜國輝
Johannes K. Chiang
張永承
Jhang, Yong-Cheng關鍵詞 自組織特徵映射網路
倒傳遞類神經網路
Elman反饋式類神經網路
徑向基底函數類神經網路
台灣股價加權指數
股票指數預測
Self-Organizing Feature Map Neural Network(SOFM)
Back Propagation Neural Network(BPN)
Elman Recurrent Neural Network(Elman)
Radial Basis Function Neural Network(RBF)
Taiwan Stock Exchange Weighted Index(TWSE)
Stock Index Forecasting日期 2011 上傳時間 4-Sep-2013 17:00:46 (UTC+8) 摘要 本研究考慮重要且對台股大盤指數走勢有連動影響的因素，主要納入對台股有領頭作用的美國三大股市，那斯達克(NASDAQ)指數、道瓊工業(Dow Jones)指數、標準普爾500(S&P500)指數;其他對台股緊密連動效果的國際股票市場，香港恆生指數、上海證券綜合指數、深圳證券綜合指數、日經225指數;以及納入左右國際經濟表現的國際原油價格走勢，美國西德州原油、中東杜拜原油和歐洲北海布蘭特原油;在宏觀經濟因素方面則考量失業率、消費者物價指數、匯率、無風險利率、美國製造業重要指標的存貨/銷貨比率、影響貨幣數量甚鉅的M1B;在技術分析方面則納入多種重要的指標，心理線 (PSY) 指標、相對強弱(RSI) 指標、威廉(WMS%R) 指標、未成熟隨機(RSV) 指標、K-D隨機指標、移動平均線(MA)、乖離率(BIAS)、包寧傑%b和包寧傑帶狀寬度(BandWidth%);所有考量因素共計35項，因為納入重要因子比較多，所以完備性較高。本研究先採用的贏者全拿(Winner-Take-All) 競爭學習策略的自組織映射網路(Self-Organizing Feature Maps, SOM)，藉由將相似資料歸屬到已身的神經元萃取出關聯分類且以計算距離來衡量神經元的離散特徵，對於探索大量且高維度的非線性複雜特徵俱有優良的因素相依性投射效果，將有利於提高預測模式精準度。在線性擬合部分則結合倒傳遞(Back-Propagation, BP)、Elman反饋式和徑向基底函數類網路(Radial-Basis-Function Network, RBF)模式為指數預測輸出，並對台股加權指數隔日收盤指數進行預測和評量。而在傳統的Elman反饋式網路只在隱藏層存在反饋機制，本研究則在輸入層和隱藏層皆建立反饋機制，將儲存在輸入層和隱藏層的過去時間資訊回饋給網路未來參考。在徑向基底函數網路方面，一般選取中心聚類點採用隨機選取方式，若能有效降低中心點個數，可降低網路複雜度，本研究導入垂直最小平方法以求取誤差最小的方式強化非監督式學習選取中心點的能力，以達到網路快速收斂，提昇網路學習品質。研究資料為台股指數交易收盤價，日期自2001/1/2，至2011/10/31共2676筆資料。訓練資料自2001/1/2至2009/12/31，共2223筆;實證測試資料自2010/1/4至2011/10/31，計453個日數。主要評估指標採用平均相對誤差(AMRE)和平均絕對誤差 (AAE)。在考慮因子較多的狀況下，實證結果顯示，在先透過SOM進行因子聚類分析之後，預測因子被分成四個組別，分別再透過BP、Elman recurrent和RBF方法進行線性擬合，平均表現方面，以RBF模式下的四個群組因子表現最佳，其中RBF模式之下的群組4，其AMRE可達到0.63%，最差的AMRE則是群組1，約為1.05%;而Elman recurrent模式下的四組群組因子之ARME則介於1.01%和1.47%之間;其中預測效果表現最差則是BP模式的預測結果。顯示RBF具有絕佳的股價預測能力。最後，在未來研究建議可以運用本文獻所探討之其他數種類神經網路模式進行股價預測。
In this study, we considering the impact factors for TWSE index tendency, mainly aimed at the three major American stock markets, NASDAQ index, Dow Jones index, S&P 500, which leading the Taiwan stock market trend; the other international stock markets, such as the Hong Kong Hang-Seng Index, Shanghai Stock Exchange Composite Index, Shenzhen Stock Exchange Composite Index, NIKKEI 225 index, which have close relationship with Taiwan stock market; we also adopt the international oil price trend, such as the West Texas Intermediate Crude Oil in American, the Dubai crude oil in Middle Eastern, North Sea Brent crude oil in European, which affects international economic performance widely; On the side of macroeconomic factors, we considering the Unemployed rate, Consumer Price Index, exchange rate, riskless rate, the Inventory to Sales ratio which it is important index of American manufacturing industry, and the M1b factor which did greatly affect to currency amounts; In the part of Technical Analysis index, we adopt several important indices, such as the Psychology Line Index (PSY), Relative Strength Index (RSI), the Wechsler Memory Scale—Revised Index (WMS%R), Row Stochastic Value Index (RSV), K-D Stochastics Index, Moving Average Line (MA), BIAS, Bollinger %b (%b), Bollinger Band Width (Band Width%);All factors total of 35 which we have considered the important factor is numerous, so the integrity is high.In this study, at first we adopt the Self-Organizing Feature Maps Network which based on the Winner-Take-All competition learning strategy, Similar information by the attribution to the body of the neuron has been extracted related categories and to calculate the distance to measure the discrete characteristics of neurons, it has excellent projection effect by exploring large and complex high-dimensional non-linear characteristics for all the dependency factors , would help to improve the accuracy of prediction models, would be able to help to improve the accuracy of prediction models. The part of the curve fitting combine with the back-propagation (Back-Propagation, BP), Elman recurrent model and radial basis function network (Radial-Basis-Function Network, RBF) model for the index prediction outputs, forecast and assessment the next close price of Taiwan stocks weighted index. In the traditional Elman recurrent network exists only one feedback mechanism in the hidden layer, in this study in the input and hidden layer feedback mechanisms are established, the previous information will be stored in the input and hidden layer and will be back to the network for future reference. In the radial basis function network, the general method is to selecting cluster center points by random selection, if we have the effectively way to reduce the number of the center points, which can reduces network complexity, in this study introduce the Orthogonal Least Squares method in order to obtain the smallest way to strengthen unsupervised learning center points selecting ability, in order to achieve convergence of the network fast, and improve network learning quality.Research data for the Trading close price of Taiwan Stock Index, the date since January 2, 2001 until September 30, 2011, total data number of 2656. since January 2, 2001 to December 31, 2009 a total number of 2223 trading close price as training data; empirical testing data, from January 4, 2010 to September 30, 2011, a total number of 433. The primary evaluation criteria adopt the Average Mean Relative Error (AMRE) and the Average Absolute Error (AAE). In the condition for consider more factors, the empirical results show that, by first through SOM for factor clustering analysis, the prediction factors were divided into four categories and then through BP, Elman recurrent and RBF methods for curve fitting, at the average performance , the four group factors of the RBF models get the best performance, the group 4 of the RBF model, the AMRE can reach 0.63%, the worst AMRE is group 1, about 1.05%; and the four groups of Elman recurrent model of ARME is between 1.01% and 1.47%; the worst prediction model is BP method. RBF has shown excellent predictive ability for stocks index. Finally, the proposal can be used in future studies of the literatures that we have explore several other methods of neural network model for stock trend forecasting.參考文獻 參考文獻[中央大學管理學院ERP中心,2010] 中央大學管理學院ERP中心,” 商業智慧”,中壢, 滄海書局2010.[余世昌, 2002] 「台灣貨幣政策指標之研究」,政治大學行政管理碩士學程論文, 2002[吳津苗, 2008] 「台灣、美國總經月數據與台股股價指數之關聯性」,中央大學產業經濟研究所碩士論文, 2008[蔡曉玲, 1993] 「台灣地區貨幣供給、匯率、分類股價因果關係實證分析」,淡江大學金融研究所碩士論文, 1993[邱建良, 1998] 「油價、貨幣供給及利率差對實質變數之影響」,淡江大學金融研究所碩士論文, 1993[鄭婉秀,吳佩珊,陳君達,陳玉瓏, 2005] 「貨幣政策、匯率與股價關連性之探討：GARCH-IRF模型之應用」,朝陽商管評論 4(2), p.73-92, 2005[聶建中, 2005] 「美國油價期貨報酬與股市報酬率之非線性關係」,淡江大學金融研究所碩士論文, 2005[陳芝瑋, 2009] 「原油價格與國際主要股價相依結構之研究」,臺灣海洋大學應用經濟研究所論文, 2009[黃姿穎, 2009] 「油價、金價、匯率與國際股市之關聯性研究」,義守大學財務金融所論文, 2009[鄭鳳媚, 2010] 「國際油價波動下美股對臺股的非線性平滑移轉關係探討」,淡江大學金融研究所碩士論文, 2010[張尹華, 2008] 「油價衝擊與股市績效:國際資金流動之跨國分析」,東海大學財務金融學所碩士論文, 2008[田宸瑄, 2007] 「國際油價、股市與景氣循環之相關分析-馬可夫轉換向量誤差修正模型的運用」,世新大學財務金融學系碩士論文, 2007[姜林杰佑, 2009] 姜林杰佑, “程式交易-觀念、方法、技術與解決方案”,二版,台北, 新陸書局2009.[Kohonen, 2001] Teuvo Kohonen, “Self-organizing map” Springer,2001. [Kohonen, Somervuo, 2002] Teuvo Kohonen, Panu Somervuo, “How to make large self-organizing maps for nonvectorial data” Neural Networks, Vol.15, pp.945-952, 2002.[Kosko, 1991] Bart Kosko, "Neural Networks and Fuzzy Systems: A Dynamical Systems Approach to Machine Intelligence" Prentice Hall, pp.145-152, 1991[Kohonen, 2006] Teuvo Kohonen, “Self-organizing neural projections” Neural Networks, Vol.19, pp.723-733, 2006.[Jang, Sun, Mizutani, 1996] J.-S. R. Jang,C.-T. Sun,E. Mizutani, “Neuro-Fuzzy and soft computing: A Computational Approach to Learning and Machine Intelligence” Prentice Hall, pp.301-308, 1996[Kohonen, 1998] Teuvo Kohonen, “The self-organizing map” Journal of Neurocompuging, Vol.21, pp.1-6, 1998.[Rumelhart, McClelland, PDP, 1989] David E. Rumelhart, James L. McClelland ,PDP Research Group "Parallel Distributed Processing" MIT press, Vol.1,pp.444-459, 1989.[Elman, 1990] Jeffrey L. Elman (1990). Finding structure in time. Cognitive Science, Vol.14, pp.79-211, 1990.[Pearlmutter, 1990] Barak A. Pearlmutter (1990). Dynamic Recurrent Neural Networks. CMU-CS-88191, Carnegie Mellon University, 1990.[Psent, Llut, 1996] D. T. Psent, X. Llut. "Training of Elman networks and dynamic system modeling" International Journal of Systents Science, Vol.27, pp.221-226, 1996.[Park, Sandberg, 1991] Jooyoung Park, Irwin W. Sandberg "Universal approximation using radial-basis-function networks" Journal Neural Computation, Vol.3, Issue 2, pp.246-257, 1991.[Park, Sandberg, 1993] Jooyoung Park, Irwin W. Sandberg "Approximation and radial-basis-function networks" Journal Neural Computation, Vol.5, Issue 2, pp.305-316, 1993.[Specht, 1990] Donald F. Specht, “Probabilistic neural networks” Neural Networks Journal, Vol.3, pp.109-118, 1990.[Specht, 1991] Donald F. Specht, “A general regression neural network” Neural Networks Journal, Vol.2, pp.568-576, 1991.[Cortes, Vapnik, 1995] Corinna Cortes Vladimir Vapnik, “Support-Vector networks” Machine Learning, pp.273-297, 1995.[Yao, Tan, Poh, 1999] Jingtao Yao, Chew-Lim Tan, Hean-Lee Poh, “Neural Networks for technical Analysis: A Study on KLCI”International Journal of Theoretical and Applied Finance,Vol.2, pp.221-241, 1999.[Tsaiha, Hsub, Laia, 1998] Ray Tsaiha, Yenshan Hsub, Charles C. Laia, “Forecasting S&P 500 stock index futures with a hybrid AI system”Decision Support Systems, Vol.23,pp.161-174, 1998.[Tsang, Ng, Kwan, Mak, Choy, 2007] Philip M. Tsang, Sin-Chun Ng, Reggie Kwan, Jacky Mak, Sheung-On Choy, “An empirical examination of the use of NN5 for Hong Kong stock price forecasting”Journal of Electronic Finance, Vol.1, pp.373-388, 2007.[Tsang, Ng, Kwan, Mak, Choy, 2007] Philip M. Tsanga, Paul Kwoka, S.O. Choya, Reggie Kwanb, S.C. Nga, Jacky Maka, Jonathan Tsangc, Kai Koongd, Tak-Lam Wonge “Design and implementation of NN5 for Hong Kong stock price forecasting” Engineering Applications of Artificial Intelligence, Vol.20, pp.453-461, 2007.[Jang, Lai, Jiang, Parng, Chien, 2004] Gia-Shuh Jang, Feipei Lai, Bor-Wei Jiang, Tai-Ming Parng, Li-Hua Chien, "Intelligent stock trading system with price trend prediction and reversal recognition using dual-module neural networks" Applied Intelligence, Vol. 3, pp.225-248, 2004.[Huang, Nakamori, Wang, 2005] Wei Huang, Yoshiteru Nakamori, Shou-Yang Wang, "Forecasting stock market movement direction with support vector machine" Computers & Operations Research, Vol.32, pp.2513-2522, 2005.[Simila, Laine, 2005] Simila, T., & Laine, S., Visual approach to supervised variableselection by self-organizing map. International Journal of Neural Systems, 15(1-2), 101-110, 2005. [Laine, Simila, 2004] Laine, S., & Simila, T. , Using SOM-based data binning tosupport supervised variable selection. In Pal, N. R., Kasabov, N.,Mudi, R. K., Pal, S. & Parui, S. K. (Eds.), Neural InformationProcessing (Vol. 3316, pp. 172-180). Berlin, 2004.[Bollinger, 2001] John A. Bollinger, "Bollinger on Bollinger Bands" McGraw-Hill, 2001.[Mathworks] MathWorks - MATLAB and Simulink for Technical Computinghttp://www.mathworks.com/help/toolbox/nnet/ug/bss36ea-1.html數據來源[1] 中央銀行全球資訊網: http://www.cbc.gov.tw/ct.asp?xItem=26178&CtNode=532&mp=1[2] 行政院主計處:http://www.dgbas.gov.tw/ct.asp?xItem=393&CtNode=2850&mp=1[3] 中華民國統計資訊網 - PC-AXIS總體經濟資料庫:http://61.60.106.82/pxweb/Dialog/statfile9L.asp[4] 中華民國經濟部(Ministry of Economic Affairs, R.O.C.)全球資訊網http://www.moea.gov.tw/Mns/populace/home/Home.aspx[5] 台灣證券交易所 - 公開資訊觀測站http://newmopsov.twse.com.tw/[6] Yahoo! Financehttp://finance.yahoo.com/[7] 鉅亨網 - 經濟指標預告_金融中心http://www.cnyes.com/economy/indicator/Page/schedule.aspx[8] MoneyCafe.comhttp://www.moneycafe.com/library/fedfundsrate.htm[9] 美國商務部(United States Department of Commerce)http://www.commerce.gov/ 描述 碩士
國立政治大學
資訊管理研究所
98356034
100資料來源 http://thesis.lib.nccu.edu.tw/record/#G0098356034 資料類型 thesis dc.contributor.advisor 姜國輝 zh_TW dc.contributor.advisor Johannes K. Chiang en_US dc.contributor.author (Authors) 張永承 zh_TW dc.contributor.author (Authors) Jhang, Yong-Cheng en_US dc.creator (作者) 張永承 zh_TW dc.creator (作者) Jhang, Yong-Cheng en_US dc.date (日期) 2011 en_US dc.date.accessioned 4-Sep-2013 17:00:46 (UTC+8) - dc.date.available 4-Sep-2013 17:00:46 (UTC+8) - dc.date.issued (上傳時間) 4-Sep-2013 17:00:46 (UTC+8) - dc.identifier (Other Identifiers) G0098356034 en_US dc.identifier.uri (URI) http://nccur.lib.nccu.edu.tw/handle/140.119/60224 - dc.description (描述) 碩士 zh_TW dc.description (描述) 國立政治大學 zh_TW dc.description (描述) 資訊管理研究所 zh_TW dc.description (描述) 98356034 zh_TW dc.description (描述) 100 zh_TW dc.description.abstract (摘要) 本研究考慮重要且對台股大盤指數走勢有連動影響的因素，主要納入對台股有領頭作用的美國三大股市，那斯達克(NASDAQ)指數、道瓊工業(Dow Jones)指數、標準普爾500(S&P500)指數;其他對台股緊密連動效果的國際股票市場，香港恆生指數、上海證券綜合指數、深圳證券綜合指數、日經225指數;以及納入左右國際經濟表現的國際原油價格走勢，美國西德州原油、中東杜拜原油和歐洲北海布蘭特原油;在宏觀經濟因素方面則考量失業率、消費者物價指數、匯率、無風險利率、美國製造業重要指標的存貨/銷貨比率、影響貨幣數量甚鉅的M1B;在技術分析方面則納入多種重要的指標，心理線 (PSY) 指標、相對強弱(RSI) 指標、威廉(WMS%R) 指標、未成熟隨機(RSV) 指標、K-D隨機指標、移動平均線(MA)、乖離率(BIAS)、包寧傑%b和包寧傑帶狀寬度(BandWidth%);所有考量因素共計35項，因為納入重要因子比較多，所以完備性較高。本研究先採用的贏者全拿(Winner-Take-All) 競爭學習策略的自組織映射網路(Self-Organizing Feature Maps, SOM)，藉由將相似資料歸屬到已身的神經元萃取出關聯分類且以計算距離來衡量神經元的離散特徵，對於探索大量且高維度的非線性複雜特徵俱有優良的因素相依性投射效果，將有利於提高預測模式精準度。在線性擬合部分則結合倒傳遞(Back-Propagation, BP)、Elman反饋式和徑向基底函數類網路(Radial-Basis-Function Network, RBF)模式為指數預測輸出，並對台股加權指數隔日收盤指數進行預測和評量。而在傳統的Elman反饋式網路只在隱藏層存在反饋機制，本研究則在輸入層和隱藏層皆建立反饋機制，將儲存在輸入層和隱藏層的過去時間資訊回饋給網路未來參考。在徑向基底函數網路方面，一般選取中心聚類點採用隨機選取方式，若能有效降低中心點個數，可降低網路複雜度，本研究導入垂直最小平方法以求取誤差最小的方式強化非監督式學習選取中心點的能力，以達到網路快速收斂，提昇網路學習品質。研究資料為台股指數交易收盤價，日期自2001/1/2，至2011/10/31共2676筆資料。訓練資料自2001/1/2至2009/12/31，共2223筆;實證測試資料自2010/1/4至2011/10/31，計453個日數。主要評估指標採用平均相對誤差(AMRE)和平均絕對誤差 (AAE)。在考慮因子較多的狀況下，實證結果顯示，在先透過SOM進行因子聚類分析之後，預測因子被分成四個組別，分別再透過BP、Elman recurrent和RBF方法進行線性擬合，平均表現方面，以RBF模式下的四個群組因子表現最佳，其中RBF模式之下的群組4，其AMRE可達到0.63%，最差的AMRE則是群組1，約為1.05%;而Elman recurrent模式下的四組群組因子之ARME則介於1.01%和1.47%之間;其中預測效果表現最差則是BP模式的預測結果。顯示RBF具有絕佳的股價預測能力。最後，在未來研究建議可以運用本文獻所探討之其他數種類神經網路模式進行股價預測。 zh_TW dc.description.abstract (摘要) In this study, we considering the impact factors for TWSE index tendency, mainly aimed at the three major American stock markets, NASDAQ index, Dow Jones index, S&P 500, which leading the Taiwan stock market trend; the other international stock markets, such as the Hong Kong Hang-Seng Index, Shanghai Stock Exchange Composite Index, Shenzhen Stock Exchange Composite Index, NIKKEI 225 index, which have close relationship with Taiwan stock market; we also adopt the international oil price trend, such as the West Texas Intermediate Crude Oil in American, the Dubai crude oil in Middle Eastern, North Sea Brent crude oil in European, which affects international economic performance widely; On the side of macroeconomic factors, we considering the Unemployed rate, Consumer Price Index, exchange rate, riskless rate, the Inventory to Sales ratio which it is important index of American manufacturing industry, and the M1b factor which did greatly affect to currency amounts; In the part of Technical Analysis index, we adopt several important indices, such as the Psychology Line Index (PSY), Relative Strength Index (RSI), the Wechsler Memory Scale—Revised Index (WMS%R), Row Stochastic Value Index (RSV), K-D Stochastics Index, Moving Average Line (MA), BIAS, Bollinger %b (%b), Bollinger Band Width (Band Width%);All factors total of 35 which we have considered the important factor is numerous, so the integrity is high.In this study, at first we adopt the Self-Organizing Feature Maps Network which based on the Winner-Take-All competition learning strategy, Similar information by the attribution to the body of the neuron has been extracted related categories and to calculate the distance to measure the discrete characteristics of neurons, it has excellent projection effect by exploring large and complex high-dimensional non-linear characteristics for all the dependency factors , would help to improve the accuracy of prediction models, would be able to help to improve the accuracy of prediction models. The part of the curve fitting combine with the back-propagation (Back-Propagation, BP), Elman recurrent model and radial basis function network (Radial-Basis-Function Network, RBF) model for the index prediction outputs, forecast and assessment the next close price of Taiwan stocks weighted index. In the traditional Elman recurrent network exists only one feedback mechanism in the hidden layer, in this study in the input and hidden layer feedback mechanisms are established, the previous information will be stored in the input and hidden layer and will be back to the network for future reference. In the radial basis function network, the general method is to selecting cluster center points by random selection, if we have the effectively way to reduce the number of the center points, which can reduces network complexity, in this study introduce the Orthogonal Least Squares method in order to obtain the smallest way to strengthen unsupervised learning center points selecting ability, in order to achieve convergence of the network fast, and improve network learning quality.Research data for the Trading close price of Taiwan Stock Index, the date since January 2, 2001 until September 30, 2011, total data number of 2656. since January 2, 2001 to December 31, 2009 a total number of 2223 trading close price as training data; empirical testing data, from January 4, 2010 to September 30, 2011, a total number of 433. The primary evaluation criteria adopt the Average Mean Relative Error (AMRE) and the Average Absolute Error (AAE). In the condition for consider more factors, the empirical results show that, by first through SOM for factor clustering analysis, the prediction factors were divided into four categories and then through BP, Elman recurrent and RBF methods for curve fitting, at the average performance , the four group factors of the RBF models get the best performance, the group 4 of the RBF model, the AMRE can reach 0.63%, the worst AMRE is group 1, about 1.05%; and the four groups of Elman recurrent model of ARME is between 1.01% and 1.47%; the worst prediction model is BP method. RBF has shown excellent predictive ability for stocks index. Finally, the proposal can be used in future studies of the literatures that we have explore several other methods of neural network model for stock trend forecasting. en_US dc.description.tableofcontents 摘要 iiAbstract iii1. 緒論 11-1 研究背景與動機 11-2 研究目的與架構 22. 文獻探討 42-1 自組織神經網路 42-1-1 Kohonen競爭式學習 52-1-2 競爭式贏家全拿(Winner-take-all)學習策略 82-1-3 內星學習(Instar learning)規則與Kohonen學習規則 92-1-4 自組織特徵映射網路(Self-Organizing Feature Maps, SOFM) 102-1-5 自組織特徵映射網路學習演算法 132-2 倒傳遞類神經網路(Back-Propagation Network) 142-3 Elman反饋式類神經網路 152-3-1 動態反饋與Elman反饋類神經網路架構 152-3-2 Elman反饋式學習 162-3-3 Elman反饋式網路演算法 162-4 徑向基底函數類神經網路(Radial-Basis-Function Network) 182-4-1 徑向基底函數架構 182-4-2 徑向基底函數演算法 192-5 其他類神經網路方法與支撐向量機器 202-5-1 機率類神經網路(Probabilistic Neural Network) 202-5-2 廣義迴歸類神經網路(General Regression Neural Network) 202-5-3 學習向量量化網路(Learning Vector Quantization) 212-5-4 支撐向量機器(Support-Vector Machine) 212-6 股價影響因素選取與應用類神經演算法於股價預測之探討 222-7 重要隔日波段技術分析指標之探討 243. 研究設計 263-1 預測模式設計 263-2 測試結果與評估 354. 實證結果與討論 364-1 聚類結果與模式 364-2 BP模式之4組群組因子預測結果 414-3 Elman反饋式模式之4組群組因子預測結果 484-4 RBF模式之4組群組因子預測結果 544-5 BP、Elman recurrent、RBF三種模式之各個群組預測結果比較 615. 結論與建議 635-1 結論 635-2 建議與未來研究方向 64參考文獻 65 zh_TW dc.format.extent 2667874 bytes - dc.format.mimetype application/pdf - dc.language.iso en_US - dc.source.uri (資料來源) http://thesis.lib.nccu.edu.tw/record/#G0098356034 en_US dc.subject (關鍵詞) 自組織特徵映射網路 zh_TW dc.subject (關鍵詞) 倒傳遞類神經網路 zh_TW dc.subject (關鍵詞) Elman反饋式類神經網路 zh_TW dc.subject (關鍵詞) 徑向基底函數類神經網路 zh_TW dc.subject (關鍵詞) 台灣股價加權指數 zh_TW dc.subject (關鍵詞) 股票指數預測 zh_TW dc.subject (關鍵詞) Self-Organizing Feature Map Neural Network(SOFM) en_US dc.subject (關鍵詞) Back Propagation Neural Network(BPN) en_US dc.subject (關鍵詞) Elman Recurrent Neural Network(Elman) en_US dc.subject (關鍵詞) Radial Basis Function Neural Network(RBF) en_US dc.subject (關鍵詞) Taiwan Stock Exchange Weighted Index(TWSE) en_US dc.subject (關鍵詞) Stock Index Forecasting en_US dc.title (題名) 應用類神經網路方法於金融時間序列預測之研究--以TWSE台股指數為例 zh_TW dc.title (題名) Using Neural Network approaches to predict financial time series research--The example of TWSE index prediction en_US dc.type (資料類型) thesis en dc.relation.reference (參考文獻) 參考文獻[中央大學管理學院ERP中心,2010] 中央大學管理學院ERP中心,” 商業智慧”,中壢, 滄海書局2010.[余世昌, 2002] 「台灣貨幣政策指標之研究」,政治大學行政管理碩士學程論文, 2002[吳津苗, 2008] 「台灣、美國總經月數據與台股股價指數之關聯性」,中央大學產業經濟研究所碩士論文, 2008[蔡曉玲, 1993] 「台灣地區貨幣供給、匯率、分類股價因果關係實證分析」,淡江大學金融研究所碩士論文, 1993[邱建良, 1998] 「油價、貨幣供給及利率差對實質變數之影響」,淡江大學金融研究所碩士論文, 1993[鄭婉秀,吳佩珊,陳君達,陳玉瓏, 2005] 「貨幣政策、匯率與股價關連性之探討：GARCH-IRF模型之應用」,朝陽商管評論 4(2), p.73-92, 2005[聶建中, 2005] 「美國油價期貨報酬與股市報酬率之非線性關係」,淡江大學金融研究所碩士論文, 2005[陳芝瑋, 2009] 「原油價格與國際主要股價相依結構之研究」,臺灣海洋大學應用經濟研究所論文, 2009[黃姿穎, 2009] 「油價、金價、匯率與國際股市之關聯性研究」,義守大學財務金融所論文, 2009[鄭鳳媚, 2010] 「國際油價波動下美股對臺股的非線性平滑移轉關係探討」,淡江大學金融研究所碩士論文, 2010[張尹華, 2008] 「油價衝擊與股市績效:國際資金流動之跨國分析」,東海大學財務金融學所碩士論文, 2008[田宸瑄, 2007] 「國際油價、股市與景氣循環之相關分析-馬可夫轉換向量誤差修正模型的運用」,世新大學財務金融學系碩士論文, 2007[姜林杰佑, 2009] 姜林杰佑, “程式交易-觀念、方法、技術與解決方案”,二版,台北, 新陸書局2009.[Kohonen, 2001] Teuvo Kohonen, “Self-organizing map” Springer,2001. [Kohonen, Somervuo, 2002] Teuvo Kohonen, Panu Somervuo, “How to make large self-organizing maps for nonvectorial data” Neural Networks, Vol.15, pp.945-952, 2002.[Kosko, 1991] Bart Kosko, "Neural Networks and Fuzzy Systems: A Dynamical Systems Approach to Machine Intelligence" Prentice Hall, pp.145-152, 1991[Kohonen, 2006] Teuvo Kohonen, “Self-organizing neural projections” Neural Networks, Vol.19, pp.723-733, 2006.[Jang, Sun, Mizutani, 1996] J.-S. R. Jang,C.-T. Sun,E. Mizutani, “Neuro-Fuzzy and soft computing: A Computational Approach to Learning and Machine Intelligence” Prentice Hall, pp.301-308, 1996[Kohonen, 1998] Teuvo Kohonen, “The self-organizing map” Journal of Neurocompuging, Vol.21, pp.1-6, 1998.[Rumelhart, McClelland, PDP, 1989] David E. Rumelhart, James L. McClelland ,PDP Research Group "Parallel Distributed Processing" MIT press, Vol.1,pp.444-459, 1989.[Elman, 1990] Jeffrey L. Elman (1990). Finding structure in time. Cognitive Science, Vol.14, pp.79-211, 1990.[Pearlmutter, 1990] Barak A. Pearlmutter (1990). Dynamic Recurrent Neural Networks. CMU-CS-88191, Carnegie Mellon University, 1990.[Psent, Llut, 1996] D. T. Psent, X. Llut. "Training of Elman networks and dynamic system modeling" International Journal of Systents Science, Vol.27, pp.221-226, 1996.[Park, Sandberg, 1991] Jooyoung Park, Irwin W. Sandberg "Universal approximation using radial-basis-function networks" Journal Neural Computation, Vol.3, Issue 2, pp.246-257, 1991.[Park, Sandberg, 1993] Jooyoung Park, Irwin W. Sandberg "Approximation and radial-basis-function networks" Journal Neural Computation, Vol.5, Issue 2, pp.305-316, 1993.[Specht, 1990] Donald F. Specht, “Probabilistic neural networks” Neural Networks Journal, Vol.3, pp.109-118, 1990.[Specht, 1991] Donald F. Specht, “A general regression neural network” Neural Networks Journal, Vol.2, pp.568-576, 1991.[Cortes, Vapnik, 1995] Corinna Cortes Vladimir Vapnik, “Support-Vector networks” Machine Learning, pp.273-297, 1995.[Yao, Tan, Poh, 1999] Jingtao Yao, Chew-Lim Tan, Hean-Lee Poh, “Neural Networks for technical Analysis: A Study on KLCI”International Journal of Theoretical and Applied Finance,Vol.2, pp.221-241, 1999.[Tsaiha, Hsub, Laia, 1998] Ray Tsaiha, Yenshan Hsub, Charles C. Laia, “Forecasting S&P 500 stock index futures with a hybrid AI system”Decision Support Systems, Vol.23,pp.161-174, 1998.[Tsang, Ng, Kwan, Mak, Choy, 2007] Philip M. Tsang, Sin-Chun Ng, Reggie Kwan, Jacky Mak, Sheung-On Choy, “An empirical examination of the use of NN5 for Hong Kong stock price forecasting”Journal of Electronic Finance, Vol.1, pp.373-388, 2007.[Tsang, Ng, Kwan, Mak, Choy, 2007] Philip M. Tsanga, Paul Kwoka, S.O. Choya, Reggie Kwanb, S.C. Nga, Jacky Maka, Jonathan Tsangc, Kai Koongd, Tak-Lam Wonge “Design and implementation of NN5 for Hong Kong stock price forecasting” Engineering Applications of Artificial Intelligence, Vol.20, pp.453-461, 2007.[Jang, Lai, Jiang, Parng, Chien, 2004] Gia-Shuh Jang, Feipei Lai, Bor-Wei Jiang, Tai-Ming Parng, Li-Hua Chien, "Intelligent stock trading system with price trend prediction and reversal recognition using dual-module neural networks" Applied Intelligence, Vol. 3, pp.225-248, 2004.[Huang, Nakamori, Wang, 2005] Wei Huang, Yoshiteru Nakamori, Shou-Yang Wang, "Forecasting stock market movement direction with support vector machine" Computers & Operations Research, Vol.32, pp.2513-2522, 2005.[Simila, Laine, 2005] Simila, T., & Laine, S., Visual approach to supervised variableselection by self-organizing map. International Journal of Neural Systems, 15(1-2), 101-110, 2005. [Laine, Simila, 2004] Laine, S., & Simila, T. , Using SOM-based data binning tosupport supervised variable selection. In Pal, N. R., Kasabov, N.,Mudi, R. K., Pal, S. & Parui, S. K. (Eds.), Neural InformationProcessing (Vol. 3316, pp. 172-180). Berlin, 2004.[Bollinger, 2001] John A. Bollinger, "Bollinger on Bollinger Bands" McGraw-Hill, 2001.[Mathworks] MathWorks - MATLAB and Simulink for Technical Computinghttp://www.mathworks.com/help/toolbox/nnet/ug/bss36ea-1.html數據來源[1] 中央銀行全球資訊網: http://www.cbc.gov.tw/ct.asp?xItem=26178&CtNode=532&mp=1[2] 行政院主計處:http://www.dgbas.gov.tw/ct.asp?xItem=393&CtNode=2850&mp=1[3] 中華民國統計資訊網 - PC-AXIS總體經濟資料庫:http://61.60.106.82/pxweb/Dialog/statfile9L.asp[4] 中華民國經濟部(Ministry of Economic Affairs, R.O.C.)全球資訊網http://www.moea.gov.tw/Mns/populace/home/Home.aspx[5] 台灣證券交易所 - 公開資訊觀測站http://newmopsov.twse.com.tw/[6] Yahoo! Financehttp://finance.yahoo.com/[7] 鉅亨網 - 經濟指標預告_金融中心http://www.cnyes.com/economy/indicator/Page/schedule.aspx[8] MoneyCafe.comhttp://www.moneycafe.com/library/fedfundsrate.htm[9] 美國商務部(United States Department of Commerce)http://www.commerce.gov/ zh_TW
