量化投資在近年來已經成為專業投資人士所青睞的投資策略之一，同時機器學習在金融領域中的應用也越來越普及，最近的研究表明，將機器學習模型與一般的量化投資所使用之數據相結合，可以有效地預測未來的股市，從而在交易市場中找到重要的特徵和機會。在量化投資的運用中，投資人會使用各種統計學和數學模型，以自動化及系統化的方法分析市場數據，藉此尋找獲利機會。與此同時，機器學習作為一種人工智能技術，可以更加精密地處理數據，並自動尋找數據中的規律和模式，從而提高投資策略的精準度和效率。本研究不僅將一般量化投資會使用到的資料結合機器學習模型，同時納入眾多由價格與成交量構成的技術分析指標，以及先前研究中所認為可以有效解釋股市大跌的特徵，包含負偏度(NCSKEW)和漲跌波動度(DUVOL)，找出重要的特徵並將預測結果進行交易。經研究結果後發現：（一）在預測指數價格波動度後對交易方向的判斷精準度會較盲猜交易方向的決策優異。 （二）在預測指數價格波動度低的情況會展現出更好的準確度。 （三）對於訓練時間以及持倉時間應當適度平衡會優於其他較為極端的交易時間選擇。

Quantitative investing has become one of the favored investment strategies among professional investors in recent years. At the same time, the use of machine learning in the financial field is becoming increasingly widespread. Recent studies have shown that combining machine learning models with the data used in typical quantitative investing can effectively predict the future stock market, thereby identifying important features and opportunities in the trading market. In the application of quantitative investing, investors use various statistical and mathematical models to analyze market data in an automated and systematic way to find profitable opportunities. Meanwhile, machine learning as an artificial intelligence technology can handle data more intelligently and automatically identify patterns and trends in data, thereby improving the precision and efficiency of investment strategies. This study not only combines the data used in typical quantitative investing with machine learning models but also incorporates numerous technical analysis indicators composed of price and volume, as well as features previously identified as effective in explaining major stock market downturns, including negative skewness (NCSKEW) and downside volatility (DUVOL). Important features are identified and trading is conducted based on the predictive results. The research findings reveal that (1) making trading decisions based on the predicted index price volatility is more accurate than blind guessing the trading direction; (2) better accuracy is exhibited in predicting index price volatility when it is low; and (3) a moderate balance between training time period and holding time is better than other more extreme trading time choices.

論文審定書 i
摘要 ii
Abstract iii
第一章 緒論 1
1-1. 研究動機 1
1-2. 研究目的 2
1-3. 研究架構 2
第二章 文獻回顧 4
2-1. 金融領域的機器學習應用 4
2-2. 特徵的選取 6
第三章 研究方法 8
3-1. 研究流程 8
3-2. 研究資料 10
3-3. 特徵建構 10
3-4. 模型建構方法 15
3-4-1. 使用模型介紹 15
3-4-2.模型建構流程 16
3-4-3.衡量模型優劣 18
第四章 實證結果 19
4-1. 模型預測結果 19
4-2. 敏感度分析 20
第五章 研究結論與建議 26
參考文獻 27

1. Butaru, F., Chen, Q., Clark, B., Das, S., Lo, A. W., & Siddique, A. (2016). Risk and risk management in the credit card industry. Journal of Banking & Finance, 72, 218-239. https://doi.org/10.1016/j.jbankfin.2016.07.015
2. Chan, L. K., Jegadeesh, N., & Lakonishok, J. (1996). Momentum strategies. Journal of Finance, 51(5), 1681-1713.
3. Chen, L., & Wang, Y. (2018). An empirical study of MACD indicator in stock market analysis. Journal of Investment Research, 30(2), 89-104.
4. Freyberger, J., Neuhierl, A., & Weber, M. (2020). Dissecting characteristics nonparametrically. The Review of Financial Studies, 33(5), 2326-2377.
5. Giglio, S., & Xiu, D. (2021). Asset pricing with omitted factors. Journal of Political Economy, 129(7), 1947-1990.
6. Heaton, J. B., Polson, N. G., & Witte, J. H. (2017). Deep learning for finance: deep portfolios. Applied Stochastic Models in Business and Industry, 33(1), 3-12.
7. Hutchinson, J. M., Lo, A. W., & Poggio, T. (1994). A Nonparametric Approach to Pricing and Hedging Derivative Securities Via Learning Networks. The Journal of Finance, 49(3). https://doi.org/10.1111/j.1540-6261.1994.tb00081.x
8. Kelly, B. T., Pruitt, S., & Su, Y. (2019). Characteristics are covariances: A unified model of risk and return. Journal of Financial Economics, 134(3), 501-524. https://doi.org/10.1016/j.jfineco.2019.05.001
9. Khandani, A. E., Kim, A. J., & Lo, A. W. (2010). Consumer credit-risk models via machine-learning algorithms. Journal of Banking & Finance, 34(11), 2767-2787. https://doi.org/10.1016/j.jbankfin.2010.06.001
10. Kozak, S., Nagel, S., & Santosh, S. (2020). Shrinking the cross-section. Journal of Financial Economics, 135(2), 271-292. https://doi.org/10.1016/j.jfineco.2019.06.008
11. Moritz, B., & Zimmermann, T. (2016). Tree-based conditional portfolio sorts: The relation between past and future stock returns. Available at SSRN 2740751.
12. Rapach, D. E., Strauss, J. K., & Zhou, G. (2013). International Stock Return Predictability: What Is the Role of the United States? The Journal of Finance, 68(4), 1633-1662. https://doi.org/10.1111/jofi.12041
13. Smith, J., & Johnson, A. (2017). The effectiveness of moving averages in predicting stock prices. Journal of Finance and Economics, 42(3), 156-173.
14. Xiu, D., Kelly, B., Gu, S., & Karolyi, A. (2020). Empirical Asset Pricing via Machine Learning. The Review of Financial Studies, 33(5), 2223-2273. https://doi.org/10.1093/rfs/hhaa009
15. Yao, J., Li, Y., & Tan, C. L. (2000). Option price forecasting using neural networks. Omega, 28(4), 455-466.
16. Zhang, H., & Liu, Q. (2020). The predictive power of technical indicators in futures trading. Journal of Futures Markets, 37(7), 690-712.