“股市是否可以預測”是一個從古至今不論學界或業界都很感興趣的議題。本研究嘗試選擇中國滬深300指數作為股市漲跌之預測標的，並利用一般投資人即可取得之最新代表每月份的總體經濟數據(包含經濟循環變數與金融市場變數)，分別對個別變數進行轉換，將之轉換成特定的動能指標後，及套入機器學習(Machine Learning)的分類決策樹模型(Classification Decision Tree Model) ，利用轉換過的動能指標對未來一個月期間預測標的之漲跌方向進行預測。使用決策樹模型的目的是要對總經變數進行預測力的篩選，我們僅保留預測準確率大於5成的變數，最後將這些通過篩選的變數套入到人工神經網絡模型(Artificial Neural Network Model)中，並一起對目標變數進行一個月期間之漲跌方向的預測。
最後預測的結果顯示，在調控參數設定與強化模型後，利用篩選過後的總經變數對滬深300未來一個月期間之漲跌方向的預測準確率最高可達6成，這個研究對於想要使用總經數據來對股市方向進行預測的個人或機構而言，是一個值得參考的依據。

Stock forecasting has become a very popular issue since time immemorial.
This paper examines the forecasting ability between the chosen stock market
index(SHA: 000300) and macroeconomic variables. The period cover in this study is
between January 2009 to December 2017. The macroeconomics variables will first
transform into specific momentum indicator. Then we will use the Classification
Decision Tree Model for further variable selection. The variable will be selected as
“capable indicator” only if the forecasting accuracy is greater than 50%. The capable
indicators will later be added in Artificial Neural Network(ANN) Model to forecast
the monthly volatility direction of the chosen stock market index simultaneously.
After optimized the parameters setting and reinforced the model, the results shows the
back-testing forecasting accuracy is up to 60%. This empirical study indicated that it
is possible to establish meaningful insight about the relationship between
macroeconomic variables and stock market index.

目錄
論文審定書.....................................................................................................................i
中文摘要........................................................................................................................ii
英文摘要....................................................................................................................... iii
第一章 緒論..................................................................................................................1
研究動機與目的......................................................................................1
研究流程..................................................................................................1
第二章 文獻回顧..........................................................................................................2
機器學習介紹..........................................................................................2
探討總體經濟變數在中國市場所扮演的角色......................................2
第三章 研究方法..........................................................................................................4
研究主題及資料來源..............................................................................4
變數轉換與說明......................................................................................5
實證研究方法與模型選擇....................................................................14
第四章 實證結果........................................................................................................22
指標分拆................................................................................................22
初步結果................................................................................................23
優化模型................................................................................................28
優化後結果............................................................................................29
第五章 結論與建議....................................................................................................35
結論........................................................................................................35
未來研究方向........................................................................................35
第六章 參考文獻........................................................................................................37

中文部分
范辛亭(2012)。基於擇時功效的股市宏觀多因素預測模型。長江證券研究報
告。
范辛亭(2012)。基於加權最小二乘法的宏觀多因素預測模型。長江證券研究報
告。
崔曉(2008)。政府政策對證券市場影響的實證分析，西南交通大學研究生學位
論文
謝金河(無日期)。港幣的指引。民國一○八年六月九日，取
自:https://www.wealth.com.tw/home/articles/17177
英文部分
Hosseini. (2011), The Role of Macroeconomics Variable on Stock Market Index in
China and India, International Journal of Economics and Finance, 3(6), 233-243
S.B.Kotsiants (2007), Supervised Machine Learning: A Review of Classification
Techniques, Emerging Artificial Intelligence Applications in Computer Engineering,
3-24
Valukonis (2013), China’s Stock Market Trends and Their Determinants Analysis
Using Market Indices, Economics and Management, 18(4), 651-660
Xiaohui Liu & Peter Sinclair (2008), Does the linkage between stock market
performance and economic growth vary across Greater China? , Applied Economics
Letters, 15:7, 505-508
Zhao (2010), Dynamic Relationship between Exchange Rate and Stock Price:
Evidence from China, Research in International Business and Finance, 24(10), 103-
112
Zhang Qianqian(2011), The impact of International Oil Price Fluctuation on China’s
Economy, Energy Procedia (5), 1360-1364