臺灣博碩士論文加值系統

本研究結合遺傳演算法與支持向量機模型，以各類技術指標值做為輸入特徵，對台灣加權指數2000年至2014年間每個交易日，預測未來一週的漲與跌兩分類，並將預測結果用於期貨實證投資上。此外並嘗試對訓練資料使用分群方法、F分數特徵挑選方法及逐步迴歸特徵挑選方法，觀看上述方法是否能夠有效提升預測準確率與投資報酬率。
最好的預測結果模型為對訓練資料使用分群且不做特徵挑選，整體樣本預測準確率為52.66%，預測漲樣本條件下準確率為57.15%，預測跌樣本條件下準確率為48.29%。由於各模型在預測漲樣本條件下準確率較高，因此實證回溯投資只針對預測漲的訊號進場。本研究實證回溯投資考慮到在不同的停損比率、停損分位比率以及槓桿倍數對投資報酬率帶來的影響。年化報酬率最高的實證回溯投資模型為對訓練資料進行分群及用F分數法對特徵進行挑選。該模型在停損分位數為0.5的條件下，停損比率設為10%，槓桿倍數設為3倍時，年化報酬率可達23.47%。
本研究證實機器學習模型可應用於證券投資領域，利用遺傳演算法結合支持向量機模型學習各類技術指標中隱含的漲跌規律，預測股市未來一週趨勢，並依預測的訊號搭配合適的交易策略進行投資，可從市場中獲取可觀的報酬。

This paper develops a prediction model which combined genetic algorithms and support vector machine to predict the five-day-ahead direction of the TAIEX. We employ technical indicators as model input features, and the model output is upward or downward signal. To confirm the model output signal is able to make return from market, we use the output signals to do backtesting on TAIFEX between 2000 and 2014. In addition, we also try to use clustering and feature selection method to training data for improving the accuracy rate and investment return.
The best prediction model is used clustering and no feature selection method to the training data. The accuracy rate of all sample is 52.66%, and the accuracy rate of the sample on the condition of predicting upward signal is 57.15%, but the accuracy rate of the sample on the condition of predicting downward signal is only 48.29%. Because our models have the higher accuracy rate on the condition of predicting upward signal, we only use upward signal to do backtesting on TAIFEX. In the backtesting, we consider the impact of stop loss quantile, stop loss rate, and leverages on the investment return. The best backtesting result of the model is used clustering and F-score feature selection method to the training data, its annualized rate of return can reach 23.47% when the the stop loss quantile sets to 0.5, stop loss rate sets to 10% and leverage sets to 3 times.
This study demonstrated that machine learning model can be used in securities investment. It can obtain abundant reward from market through our method.

論文審定書 i
誌謝辭 ii
摘要 iii
Abstract iv
第一章 緒論 1
1.1 研究背景與動機 1
1.2 研究目的與問題 2
1.3 研究架構與流程 2
第二章 文獻回顧 4
2.1 支持向量機 4
2.1.1 資料線性可分隔支持向量機 5
2.1.2 資料線性不可分隔支持向量機 8
2.1.3 非線性支持向量機 9
2.2 遺傳演算法 11
2.3 特徵挑選方法 11
2.3.1 F分數法(F-score) 11
2.3.2 逐步迴歸法(Stepwise Regression) 12
2.4 支持向量機在證券投資的應用 12
2.4.1 國外文獻 13
2.4.2 國內文獻 14
第三章 研究方法 16
3.1 研究模型 16
3.2 預測分類 17
3.3 分群方法 18
3.4 訓練期與預測期選擇方法 19
3.5 特徵資料 19
3.5.1 移動平均線(Moving Average; MA) 20
3.5.2 乖離率(BIAS) 20
3.5.3 指數平滑異同移動平均線(Moving Average Convergence and Divergence; MACD) 21
3.5.4 相對強弱指標(Relative Strength Index; RSI) 21
3.5.5 KD隨機指標(Stochastics Line; KD) 22
3.5.6 威廉指標(Williams %R) 22
3.5.7 趨向指標(Directional Movement Index; DMI) 23
3.5.8 動量指標(Momentum Index; MTM) 25
3.5.9 移動平均量(Moving Average Volume; MAV) 25
3.5.10 量強弱指標(Volume Ratio; VR) 27
3.5.11 人氣指標(AR)與意願指標(BR) 27
3.6 資料正規化 31
3.7 特徵挑選方法 31
3.7.1 F分數法(F-score) 32
3.7.2 逐步迴歸法(Stepwise Regression) 32
3.8 GA-SVM模型 32
第四章 模型與實證結果與分析 36
4.1 資料來源 36
4.2 使用軟體 36
4.3 模型預測 36
4.3.1 模型設定 36
4.3.2 模型預測結果 37
4.4 實證回溯投資 42
4.4.1 投資規則設定 42
4.4.2 實證投資結果 42
第五章 結論與建議 52
5.1 研究結論 52
參考文獻 54

一、中文文獻
池庭志(2008)，利用支持向量機來預測股市，中華大學資訊管理所，碩士論文。
林蔚宗(2011)，應用平滑支撐向量機於台幣黃金期貨的投資策略，國立台灣科技大學資訊管理所，碩士論文。
曹鈞斐(2007)，運用平滑支撐向量機於期貨市場擇時策略之研究，國立台灣科技大學資訊管理所，碩士論文。
陳寬裕(2006)，結合遺傳演算法與支援向量迴歸於台灣股票加權指數之預測，計量管理期刊，第三卷，第一期，1-18頁。
鄭敦維(2012)，一個基植於遺傳演算法與模糊理論最佳化之支援向量機選股模型，國立高雄大學資訊工程研究所，碩士論文。
二、英文文獻
Boser, B. E., Guyon, I. M., & Vapnik, V. N. (1992, July). A training algorithm for optimal margin classifiers. In Proceedings of the fifth annual workshop on Computational learning theory (pp. 144-152). ACM.
Chang, C. C., & Lin, C. J. (2011). LIBSVM: A library for support vector machines. ACM Transactions on Intelligent Systems and Technology (TIST),2(3), 27.
Chen, Y. W., & Lin, C. J. (2006). Combining SVMs with various feature selection strategies. In Feature extraction (pp. 315-324). Springer Berlin Heidelberg.
Chipperfield, A. J., & Fleming, P. J. (1995, January). The MATLAB genetic algorithm toolbox. In Applied control techniques using MATLAB, IEE Colloquium on (pp. 10-1). IET.
Cortes, C., & Vapnik, V. (1995). Support-vector networks. Machine learning,20(3), 273-297.
Holland, J. H. (1975). Adaptation in natural and artificial system: an introduction with application to biology, control and artificial intelligence. Ann Arbor, University of Michigan Press.
Hsu, C. W., Chang, C. C., & Lin, C. J. (2003). A practical guide to support vector classification.
Huang, W., Nakamori, Y., & Wang, S. Y. (2005). Forecasting stock market movement direction with support vector machine. Computers & Operations Research, 32(10), 2513-2522.
Kara, Y., Boyacioglu, M. A., & Baykan, Ö. K. (2011). Predicting direction of stock price index movement using artificial neural networks and support vector machines: The sample of the Istanbul Stock Exchange. Expert systems with Applications, 38(5), 5311-5319.
Kim, K. J. (2003). Financial time series forecasting using support vector machines. Neurocomputing, 55(1), 307-319.
Lee, M. C. (2009). Using support vector machine with a hybrid feature selection method to the stock trend prediction. Expert Systems with Applications, 36(8), 10896-10904.
SchiilkopP, B., Burgest, C., & Vapnik, V. (1995). Extracting support data for a given task. no. x.
Tay, F. E., & Cao, L. (2001). Application of support vector machines in financial time series forecasting. Omega, 29(4), 309-317.