簡易檢索 / 詳目顯示

研究生: 林靖荃
Lin, Jing-Cyuan
論文名稱: 利用鬆弛變量核密度估計方法預測股票指數
Forecasting stock indices with relaxed variable kernel density estimation
指導教授: 張天豪
Chang, Tien-Hao
共同指導教授: 劉裕宏
Liu, Yu-hong
學位類別: 碩士
Master
系所名稱: 電機資訊學院 - 電機工程學系
Department of Electrical Engineering
論文出版年: 2014
畢業學年度: 102
語文別: 中文
論文頁數: 30
中文關鍵詞: 上海綜合指數標準普爾500指數日經225指數鬆弛變量核密度估計方法支援向量機類神經網路
外文關鍵詞: Shanghai Composite Index, S&P 500 Index, Nikkei 225 Index, Relaxed Variable Kernel Density Estimator, Support Vector Machine, Neural Network
相關次數: 點閱:116下載:0
分享至:
查詢本校圖書館目錄 查詢臺灣博碩士論文知識加值系統 勘誤回報
  • 股票價格的時間序列具有動態、非線性和混沌性質。股票市場受多種因素的影響,預測股價或股價指數,因為受到多種因素的影響,通常會有很大的誤差。在本文中,我們預測上海綜合指數(Shanghai Composite Index, SCI)、標準普爾500指數(Standard & Poor's 500 Index, S&P 500)與日經225指數(Nikkei 225 Index),從1993年1月至2009年12月的每月收盤價格。
    我們發現股價長期走勢明顯時,將預測目標從上海綜合指數改為上海綜合指數變化百分比,會有明顯較好的預測效果。當特徵向量加入中國的三大主要出口國的股價指數變化百分比時,上海綜合指數變化百分比預測效果略為進步。當使用鬆弛變量核密度估計方法(relaxed variable kernel density estimator, RVKDE)時,相較支援向量機(support vector machine, SVM)與類神經網路(neural network)預測效果較佳。
    我們同時做強健度測試,將預測目標從標準普爾500指數改為標準普爾500指數變化百分比後,預測效果明顯進步。當使用鬆弛變量核密度估計方法(relaxed variable kernel density estimator, RVKDE)時,相較支援向量機(support vector machine, SVM)與類神經網路(neural network)預測效果較佳。
    將預測目標從日經225指數改為日經225指數變化百分比後,預測效果明顯進步。當使用鬆弛變量核密度估計方法(relaxed variable kernel density estimator, RVKDE)時,相較支援向量機(support vector machine, SVM)與類神經網路(neural network)預測效果較佳。

    Stock prices is time series. Stock prices are basically dynamic, non-linear, and chaotic in stock markets. Stock markets are influenced by many factors. Predicting stock price or stock index with the noisy data directly is usually subject to large errors. In this research, we try to predict the monthly closing price data with Shanghai Composite Index, Standard & Poor's 500 Index and Nikkei 225 Index from January 1993 to December 2009.
    We discover that while the tendency of the stock index price is obvious, turning the Shanghai Composite Index price of the predicted target into percentage change in Shanghai Composite Index price has the better prediction. We also do robustness test. We forecast S&P 500 Index closing price and Nikkei 225 Index closing price.
    We also find that the prediction of using the RVKDE (relaxed variable kernel density estimator) has the better accuracy than SVM (support vector machine) and neural network. While adding the percentage change of the stock index of the three main export countries for target country, the prediction is also improved.

    目錄.......................1 表目錄.....................3 圖目錄.....................4 第一章 緒論.................5 第二章 相關研究..............7 2.1 上海綜合指數.............7 2.2 標準普爾500指數..........9 2.3 日經225指數............10 2.4 機器學習技術............10 2.4.1鬆弛變量核密度估計方法...11 2.4.2支援向量機.............11 2.4.3比較兩個分類器差異.......13 第三章 資料收集與方法.........15 3.1 資料收集................15 3.1.1 資料庫................15 3.1.2 資料處理..............17 3.2 實驗方法設計.............19 第四章 實驗結果與討論分析.......20 4.1 實驗流程.................20 4.1.1 步驟一:資料切割........21 4.1.2 步驟二:尺度化..........21 4.1.3 步驟三:參數調整.........22 4.2 效能評估準則..............22 4.2.1 MAE...................22 4.2.2 RMSE..................22 4.3 與其他實驗設計比較..........23 第五章 結論與未來展望...........29 5.1 結論.....................29 5.2 未來展望..................29 參考文獻......................30

    1.Kyong Joo Oh and Kyoung-jae Kim: Analyzing stock market tick data using piecewise nonlinear model. 2002: Expert Systems with Applications.
    2.Philip Hans Franses and Hendrik Ghijsels: Additive outliers, GARCH and forecasting volatility. 1999: International Journal of Forecasting.
    3.Nicholas Sarantis: Nonlinearities, cyclical behaviour and predictability in stock markets: international evidence. 2001: International Journal of Forecasting.
    4.David Enke and Suraphan Thawornwong: The use of data mining and neural networks for forecasting stock market returns. 2005: Expert Systems with Applications.
    5.G. Armano, M. Marchesi, A. Murru: A hybrid genetic-neural architecture for stock indexes forecasting. 2005: Information Sciences.
    6.Y. Oyang ea: Data classification with a relaxed model of variable kernel density estimation. pp 2005:2831-2836.
    7.Vapnik CCaV: Support vector machine. pp 1995, 20:273-297.
    8.Shih C-W: Support Vector Machine. http://eeilimenckuedutw/knowledgebase/zhi-yuan-xiang-liang-ji-support-vector-machine 2011.
    9.Jian-Zhou Wang, Ju-Jie Wang, Zhe-George Zhang, Shu-Po Guo:Forecasting stock indices with back propagation neural network. 2011: Expert Systems with Applications.

    無法下載圖示 校內:2024-12-31公開
    校外:不公開
    電子論文尚未授權公開,紙本請查館藏目錄
    QR CODE