Publications-Theses
Article View/Open
Publication Export
-
題名 基於多任務遷移學習之上市公司財報基本面與產業表現關聯股價預測
Stock Price Prediction based on Financial Statement and Industry Status using Multi-task Transfer Learning作者 古昊中
Ku, Hao-Chung貢獻者 姜國輝
Chiang, Kuo-Huie
古昊中
Ku, Hao-Chung關鍵詞 機器學習
類神經網路
長短期記憶網路
遷移學習
多任務學習
股市預測
財務報表
Machine Learning
Neural Network
Long Short-Term Memory
Transfer Learning
Multi-task Learning
Stock Market Forecasting
Financial Statement日期 2019 上傳時間 5-Sep-2019 15:44:32 (UTC+8) 摘要 隨著資訊科技快速的發展,許多新的科技技術與創新應用不斷地出現,並受惠於硬體技術的大幅進步,在這資訊爆炸的年代,電腦能夠負擔技術上以及應用上的需求,為社會提供許多的便利、可靠性。同時提供給業界各個領域多元的解決方案與近破壞式的創新,讓商業不斷地進化、革新。以金融產業來說,金融業因涉及資金的流通,必須兼顧信用、安全、精準等,對於改變以及創新往往趨於保守,但因人工智慧的興起,看到了技術所帶來的好處並為上述的顧慮提供保證,開始帶動了金融科技的革命,為金融業服務提供有別於一般所設想的模式,並帶來可觀的成本降低以及獲益增加,使各個公司紛紛擁抱技術,享受技術所帶來的優勢與效益。本研究以半導體產業之上市公司為例,利用公司之資產負債表、損益表、現金流量表內會計項目作為公司股價預測之依據,藉由選取財務報表中的會計項目進行公司基本面資訊之取得。在模型方面,本研究採用類神經網路並結合長短期記憶網路作為預測之技術,並透過多任務學習的方式萃取產業基本面特徵與股價指數的潛在結構,將之應用於特定公司,以取得相應的市場價值,建構出合適於相同產業別多家上市公司之股價預測模型,實現單一模型具備多家公司預測之能力。藉由公司之間資料的輔助學習,以及公司與產業之間之鏈結,減少因資料涵蓋面不足所限制的預測效果以及增加現實中公司與產業間氣氛、趨勢實際的互動與牽引。研究結果顯示,本模型之預測具備一定泛化能力,能降低模型發生之誤差,不會因特定公司資訊稀少,導致預測效果特別不佳。此外,本研究樣本資料期間為十一年,結果顯示模型預測效果對於近期的預測有顯著且穩定的效果。
With the rapid development of technology, new technologies are bringing into solutions and disruptive innovation to our society. Take financial industry as an example, the financial industry’s services and products are usually related to the circulation of funds. It results in the change and innovation tending to be conservative. However, due to the rise of artificial intelligence, companies recognized the benefits, safety and promise of technologies, and started to embrace those technologies which can provide new business model and bring benefits.This study retrieves the information from published financial statements of listed semiconductor industry in Taiwan as the basis of evaluation the stock prices. This study combines long short-term memory and neural network with multi-task learning to extract the hidden structure of industrial basic features and stock index, which then applies to specific company and gets the potential market value in proportion. The results show the capability of generalizing the prediction to other similar companies which might lack for complete financial metrics. Over a span of eleven years collected data, the results also present significant and stable performance especially for the prediction of the recent years.參考文獻 中文文獻林美花,2009,中華無形資產鑑價研究發展協會企業評價認證班第十九期講義林逢煥,1994,應用基因法則最佳化類神經網路建立財務分析模式之研究,國立交通大學工業工程研究所碩士論文林瑞瑤,1994,以財務比率建立股票投資績效預測模式,國立中山大學財務管理研究所碩士論文陳思雅,2016,大數據在保險業的應用:以台灣壽險公司破產預測為例,國立政治大學風險管理與保險研究所碩士論文陳慧真,2009,台灣主要水泥公司之財務報表分析與企業價值衡量,國立政治大學經營管理碩士學程(EMBA)碩士論文張家瑋,2017,財務報表資訊在偵測財務危機上的有用性:個案研究,國立政治大學會計研究所碩士論文黃國裕,1994,財務比率在股票超常報酬之預測能力分析-類神經網路法,國立中央大學資訊管理研究所碩士論文游崇智,1996,應用類神經網路模擬多變量計量模式於台灣股市之分析與預測,私立中原大學企業管理研究所碩士論文游淑禎,1998,類神經網路應用於台灣股市預測,臺灣銀行季刊第四十九卷第三期,民國87年9月,27-59甄典蕙,2015,財務報表舞弊偵測模型之建立:以中國上市公司為例,國立政治大學會計研究所碩士論文蔡惠玲,2005,運用財報資訊評估企業經營績效與預測財務危機之研究:以分析損益表及現金流量表之資訊為主,國立政治大學會計研究所碩士論文劉慧敏,2001,多目標遺傳演算法於基本面選股策略之應用,國立中央大學資訊管理研究所碩士論文蘇嘉雄,2014,以財務報表資訊為台灣股票市場建構最適資產配置,國立政治大學風險管理與保險研究所碩士論文英文文獻Ball, R., & Brown, P. (1968). An empirical evaluation of accounting income numbers. Journal of accounting research, 159-178.Caruana, R. (1997). Multitask learning. Machine learning, 28(1), 41-75.Dai, W., Yang, Q., Xue, G. R., & Yu, Y. (2008, July). Self-taught clustering. In Proceedings of the 25th international conference on Machine learning (pp. 200-207). ACM.Dietterich, T. G., Pratt, L., & Thrun, S. (1997). Special issue on inductive transfer. Machine Learning, 28(1).Dorina, P., Melinda, K., & Klara, S. (2012). Contemporary approaches of company performance analysis based on relevant financial information. Annals of Faculty of Economics, 1(2), 708-715.Ganin, Y., Ustinova, E., Ajakan, H., Germain, P., Larochelle, H., Laviolette, F., ... & Lempitsky, V. (2016). Domain-adversarial training of neural networks. The Journal of Machine Learning Research, 17(1), 2096-2030.Hartigan, J. A. (1972). Direct clustering of a data matrix. Journal of the american statistical association, 67(337), 123-129.Hochreiter, S., & Schmidhuber, J. (1997). Long short-term memory. Neural computation, 9(8), 1735-1780.Hopfield, J. J. (1982). Neural networks and physical systems with emergent collective computational abilities. Proceedings of the national academy of sciences, 79(8), 2554-2558.Johnson, M., Schuster, M., Le, Q. V., Krikun, M., Wu, Y., Chen, Z., ... & Hughes, M. (2017). Google’s multilingual neural machine translation system: Enabling zero-shot translation. Transactions of the Association for Computational Linguistics, 5, 339-351.Kang, H. B. (2010). A Case Study On The Archer Daniels Midland (ADM) Company’s Financial Statement Analysis: Strengths And Weaknesses. Journal of Business Case Studies, 6(3), 65.Krizhevsky, A., Sutskever, I., & Hinton, G. E. (2012). Imagenet classification with deep convolutional neural networks. In Advances in neural information processing systems (pp. 1097-1105).Kryzanowski, L., Galler, M., & Wright, D. W. (1993). Using artificial neural networks to pick stocks. Financial Analysts Journal, 49(4), 21-27.McCarthy, J., Minsky, M. L., Rochester, N., & Shannon, C. E. (2006). A proposal for the dartmouth summer research project on artificial intelligence, august 31, 1955. AI magazine, 27(4), 12.McCulloch, W. S., & Pitts, W. (1943). A logical calculus of the ideas immanent in nervous activity. The bulletin of mathematical biophysics, 5(4), 115-133.Ou, J. A., & Penman, S. H. (1989). Financial statement analysis and the prediction of stock returns. Journal of accounting and economics, 11(4), 295-329.Pearson, K. (1901). LIII. On lines and planes of closest fit to systems of points in space. The London, Edinburgh, and Dublin Philosophical Magazine and Journal of Science, 2(11), 559-572.Raina, R., Battle, A., Lee, H., Packer, B., & Ng, A. Y. (2007, June). Self-taught learning: transfer learning from unlabeled data. In Proceedings of the 24th international conference on Machine learning (pp. 759-766). ACM.Rumelhart, D. E., Hinton, G. E., & Williams, R. J. (1988). Learning representations by back-propagating errors. Cognitive modeling, 5(3), 1.Soekarno, S., & Azhari, D. A. (2009). Analysis of Financial Ratio to Distinguish Indonesia Joint Venture General Insurance Company Performance using Discriminant Analysis. The Asian journal of technology management, 2(2), 110-122. 描述 碩士
國立政治大學
資訊管理學系
106356016資料來源 http://thesis.lib.nccu.edu.tw/record/#G0106356016 資料類型 thesis dc.contributor.advisor 姜國輝 zh_TW dc.contributor.advisor Chiang, Kuo-Huie en_US dc.contributor.author (Authors) 古昊中 zh_TW dc.contributor.author (Authors) Ku, Hao-Chung en_US dc.creator (作者) 古昊中 zh_TW dc.creator (作者) Ku, Hao-Chung en_US dc.date (日期) 2019 en_US dc.date.accessioned 5-Sep-2019 15:44:32 (UTC+8) - dc.date.available 5-Sep-2019 15:44:32 (UTC+8) - dc.date.issued (上傳時間) 5-Sep-2019 15:44:32 (UTC+8) - dc.identifier (Other Identifiers) G0106356016 en_US dc.identifier.uri (URI) http://nccur.lib.nccu.edu.tw/handle/140.119/125528 - dc.description (描述) 碩士 zh_TW dc.description (描述) 國立政治大學 zh_TW dc.description (描述) 資訊管理學系 zh_TW dc.description (描述) 106356016 zh_TW dc.description.abstract (摘要) 隨著資訊科技快速的發展,許多新的科技技術與創新應用不斷地出現,並受惠於硬體技術的大幅進步,在這資訊爆炸的年代,電腦能夠負擔技術上以及應用上的需求,為社會提供許多的便利、可靠性。同時提供給業界各個領域多元的解決方案與近破壞式的創新,讓商業不斷地進化、革新。以金融產業來說,金融業因涉及資金的流通,必須兼顧信用、安全、精準等,對於改變以及創新往往趨於保守,但因人工智慧的興起,看到了技術所帶來的好處並為上述的顧慮提供保證,開始帶動了金融科技的革命,為金融業服務提供有別於一般所設想的模式,並帶來可觀的成本降低以及獲益增加,使各個公司紛紛擁抱技術,享受技術所帶來的優勢與效益。本研究以半導體產業之上市公司為例,利用公司之資產負債表、損益表、現金流量表內會計項目作為公司股價預測之依據,藉由選取財務報表中的會計項目進行公司基本面資訊之取得。在模型方面,本研究採用類神經網路並結合長短期記憶網路作為預測之技術,並透過多任務學習的方式萃取產業基本面特徵與股價指數的潛在結構,將之應用於特定公司,以取得相應的市場價值,建構出合適於相同產業別多家上市公司之股價預測模型,實現單一模型具備多家公司預測之能力。藉由公司之間資料的輔助學習,以及公司與產業之間之鏈結,減少因資料涵蓋面不足所限制的預測效果以及增加現實中公司與產業間氣氛、趨勢實際的互動與牽引。研究結果顯示,本模型之預測具備一定泛化能力,能降低模型發生之誤差,不會因特定公司資訊稀少,導致預測效果特別不佳。此外,本研究樣本資料期間為十一年,結果顯示模型預測效果對於近期的預測有顯著且穩定的效果。 zh_TW dc.description.abstract (摘要) With the rapid development of technology, new technologies are bringing into solutions and disruptive innovation to our society. Take financial industry as an example, the financial industry’s services and products are usually related to the circulation of funds. It results in the change and innovation tending to be conservative. However, due to the rise of artificial intelligence, companies recognized the benefits, safety and promise of technologies, and started to embrace those technologies which can provide new business model and bring benefits.This study retrieves the information from published financial statements of listed semiconductor industry in Taiwan as the basis of evaluation the stock prices. This study combines long short-term memory and neural network with multi-task learning to extract the hidden structure of industrial basic features and stock index, which then applies to specific company and gets the potential market value in proportion. The results show the capability of generalizing the prediction to other similar companies which might lack for complete financial metrics. Over a span of eleven years collected data, the results also present significant and stable performance especially for the prediction of the recent years. en_US dc.description.tableofcontents 第一章 緒論 1第一節 研究背景 1第二節 研究動機與目的 2第二章 文獻探討 4第一節 財務報表運用於預測之研究 4第二節 機器學習 7一、人工神經網路(Artificial Neural Network, ANN) 7二、遞歸神經網路(Recurrent Neural Network, RNN) 8三、長短期記憶網路(Long Short-Term Memory, LSTM) 9四、遷移學習(Transfer Learning) 9第三節 主成分分析 11第三章 研究方法 13第一節 資料蒐集 13第二節 資料前處理 14一、特徵指標選取 14二、各項指標平均標準 24三、補值 25第三節 系統設計 26一、多任務學習 26二、多輸出模型 27第四節 預測模型設計 28第四章 研究結果 31第一節 訓練結果 31一、股價預測與真實股價誤差 31二、主要輸出損失函數 32三、公司特徵指標損失函數 33四、產業特徵指標損失函數 34第二節 測試結果 34一、股價預測與真實股價誤差 34二、主要輸出損失函數 35三、公司特徵指標損失函數 36四、產業特徵指標損失函數 37第三節 實驗模型驗證 40一、長期(一年) 41二、中期(半年) 41三、短期(一季) 42第五章 研究結論與未來建議 43參考文獻 44中文文獻 44英文文獻 45附錄 48 zh_TW dc.format.extent 31242581 bytes - dc.format.mimetype application/pdf - dc.source.uri (資料來源) http://thesis.lib.nccu.edu.tw/record/#G0106356016 en_US dc.subject (關鍵詞) 機器學習 zh_TW dc.subject (關鍵詞) 類神經網路 zh_TW dc.subject (關鍵詞) 長短期記憶網路 zh_TW dc.subject (關鍵詞) 遷移學習 zh_TW dc.subject (關鍵詞) 多任務學習 zh_TW dc.subject (關鍵詞) 股市預測 zh_TW dc.subject (關鍵詞) 財務報表 zh_TW dc.subject (關鍵詞) Machine Learning en_US dc.subject (關鍵詞) Neural Network en_US dc.subject (關鍵詞) Long Short-Term Memory en_US dc.subject (關鍵詞) Transfer Learning en_US dc.subject (關鍵詞) Multi-task Learning en_US dc.subject (關鍵詞) Stock Market Forecasting en_US dc.subject (關鍵詞) Financial Statement en_US dc.title (題名) 基於多任務遷移學習之上市公司財報基本面與產業表現關聯股價預測 zh_TW dc.title (題名) Stock Price Prediction based on Financial Statement and Industry Status using Multi-task Transfer Learning en_US dc.type (資料類型) thesis en_US dc.relation.reference (參考文獻) 中文文獻林美花,2009,中華無形資產鑑價研究發展協會企業評價認證班第十九期講義林逢煥,1994,應用基因法則最佳化類神經網路建立財務分析模式之研究,國立交通大學工業工程研究所碩士論文林瑞瑤,1994,以財務比率建立股票投資績效預測模式,國立中山大學財務管理研究所碩士論文陳思雅,2016,大數據在保險業的應用:以台灣壽險公司破產預測為例,國立政治大學風險管理與保險研究所碩士論文陳慧真,2009,台灣主要水泥公司之財務報表分析與企業價值衡量,國立政治大學經營管理碩士學程(EMBA)碩士論文張家瑋,2017,財務報表資訊在偵測財務危機上的有用性:個案研究,國立政治大學會計研究所碩士論文黃國裕,1994,財務比率在股票超常報酬之預測能力分析-類神經網路法,國立中央大學資訊管理研究所碩士論文游崇智,1996,應用類神經網路模擬多變量計量模式於台灣股市之分析與預測,私立中原大學企業管理研究所碩士論文游淑禎,1998,類神經網路應用於台灣股市預測,臺灣銀行季刊第四十九卷第三期,民國87年9月,27-59甄典蕙,2015,財務報表舞弊偵測模型之建立:以中國上市公司為例,國立政治大學會計研究所碩士論文蔡惠玲,2005,運用財報資訊評估企業經營績效與預測財務危機之研究:以分析損益表及現金流量表之資訊為主,國立政治大學會計研究所碩士論文劉慧敏,2001,多目標遺傳演算法於基本面選股策略之應用,國立中央大學資訊管理研究所碩士論文蘇嘉雄,2014,以財務報表資訊為台灣股票市場建構最適資產配置,國立政治大學風險管理與保險研究所碩士論文英文文獻Ball, R., & Brown, P. (1968). An empirical evaluation of accounting income numbers. Journal of accounting research, 159-178.Caruana, R. (1997). Multitask learning. Machine learning, 28(1), 41-75.Dai, W., Yang, Q., Xue, G. R., & Yu, Y. (2008, July). Self-taught clustering. In Proceedings of the 25th international conference on Machine learning (pp. 200-207). ACM.Dietterich, T. G., Pratt, L., & Thrun, S. (1997). Special issue on inductive transfer. Machine Learning, 28(1).Dorina, P., Melinda, K., & Klara, S. (2012). Contemporary approaches of company performance analysis based on relevant financial information. Annals of Faculty of Economics, 1(2), 708-715.Ganin, Y., Ustinova, E., Ajakan, H., Germain, P., Larochelle, H., Laviolette, F., ... & Lempitsky, V. (2016). Domain-adversarial training of neural networks. The Journal of Machine Learning Research, 17(1), 2096-2030.Hartigan, J. A. (1972). Direct clustering of a data matrix. Journal of the american statistical association, 67(337), 123-129.Hochreiter, S., & Schmidhuber, J. (1997). Long short-term memory. Neural computation, 9(8), 1735-1780.Hopfield, J. J. (1982). Neural networks and physical systems with emergent collective computational abilities. Proceedings of the national academy of sciences, 79(8), 2554-2558.Johnson, M., Schuster, M., Le, Q. V., Krikun, M., Wu, Y., Chen, Z., ... & Hughes, M. (2017). Google’s multilingual neural machine translation system: Enabling zero-shot translation. Transactions of the Association for Computational Linguistics, 5, 339-351.Kang, H. B. (2010). A Case Study On The Archer Daniels Midland (ADM) Company’s Financial Statement Analysis: Strengths And Weaknesses. Journal of Business Case Studies, 6(3), 65.Krizhevsky, A., Sutskever, I., & Hinton, G. E. (2012). Imagenet classification with deep convolutional neural networks. In Advances in neural information processing systems (pp. 1097-1105).Kryzanowski, L., Galler, M., & Wright, D. W. (1993). Using artificial neural networks to pick stocks. Financial Analysts Journal, 49(4), 21-27.McCarthy, J., Minsky, M. L., Rochester, N., & Shannon, C. E. (2006). A proposal for the dartmouth summer research project on artificial intelligence, august 31, 1955. AI magazine, 27(4), 12.McCulloch, W. S., & Pitts, W. (1943). A logical calculus of the ideas immanent in nervous activity. The bulletin of mathematical biophysics, 5(4), 115-133.Ou, J. A., & Penman, S. H. (1989). Financial statement analysis and the prediction of stock returns. Journal of accounting and economics, 11(4), 295-329.Pearson, K. (1901). LIII. On lines and planes of closest fit to systems of points in space. The London, Edinburgh, and Dublin Philosophical Magazine and Journal of Science, 2(11), 559-572.Raina, R., Battle, A., Lee, H., Packer, B., & Ng, A. Y. (2007, June). Self-taught learning: transfer learning from unlabeled data. In Proceedings of the 24th international conference on Machine learning (pp. 759-766). ACM.Rumelhart, D. E., Hinton, G. E., & Williams, R. J. (1988). Learning representations by back-propagating errors. Cognitive modeling, 5(3), 1.Soekarno, S., & Azhari, D. A. (2009). Analysis of Financial Ratio to Distinguish Indonesia Joint Venture General Insurance Company Performance using Discriminant Analysis. The Asian journal of technology management, 2(2), 110-122. zh_TW dc.identifier.doi (DOI) 10.6814/NCCU201900737 en_US