TCN-based Futures Prediction Using Financial Indices Bargain Chips, and Forum Messages

以作者查詢圖書館館藏

、以作者查詢臺灣博碩士

、以作者查詢全國書目

、勘誤回報

、線上人數：78

、訪客IP：18.191.132.7

姓名

謝莉醇(Li-Chun Hsieh) 查詢紙本館藏

畢業系所

資訊工程學系在職專班

論文名稱

(TCN-based Futures Prediction Using Financial Indices Bargain Chips, and Forum Messages)

相關論文

★ Dynamic Overlay Construction for Mobile Target Detection in Wireless Sensor Networks	★ 車輛導航的簡易繞路策略
★ 使用傳送端電壓改善定位	★ 利用車輛分類建構車載網路上的虛擬骨幹
★ Why Topology-based Broadcast Algorithms Do Not Work Well in Heterogeneous Wireless Networks?	★ 針對移動性目標物的有效率無線感測網路
★ 適用於無線隨意網路中以關節點為基礎的分散式拓樸控制方法	★ A Review of Existing Web Frameworks
★ 將感測網路切割成貪婪區塊的分散式演算法	★ 無線網路上Range-free的距離測量
★ Inferring Floor Plan from Trajectories	★ An Indoor Collaborative Pedestrian Dead Reckoning System
★ Dynamic Content Adjustment In Mobile Ad Hoc Networks	★ 以影像為基礎的定位系統
★ 大範圍無線感測網路下分散式資料壓縮收集演算法	★ 車用WiFi網路中的碰撞分析

檔案

[Endnote RIS 格式]

[Bibtex 格式]

[相關文章]

[文章引用]

[完整記錄]

[館藏目錄]

[檢視]

[下載]

本電子論文使用權限為同意立即開放。
已達開放權限電子全文僅授權使用者為學術研究之目的，進行個人非營利性質之檢索、閱讀、列印。
請遵守中華民國著作權法之相關規定，切勿任意重製、散佈、改作、轉貼、播送，以免觸法。

摘要(中)

傳統的股票和/或期貨價格預測研究多以過去的股票和/或期貨價格和技術指標為特徵，如KD，RSI，MACD等。很少有研究將討論區留言或法人籌碼作為股票和/或期貨價格預測。在本研究中，PTT和CMoney論壇的討論區留言將使用再訓練的BERT轉換成每日情緒向量，然後以每日情緒向量和三個法人籌碼為特徵，訓練GRU和TCN模型。實驗結果表明，TCN在MAE、MAPE、RMSE和準確率方面都基於GRU的RNN模型。而轉換而成的每日情緒向量、法人籌碼和討論區留言在期貨價格預測中都被證實是有用的。基於歷史期貨價格的市場模擬表明，使用 TCN模型的簡單投資策略，利用技術指標、法人籌碼和討論區留言，可以在一年之間賺取的投資收益為成本的7倍以上。

摘要(英)

Traditional research on the stock and/or futures price prediction mostly uses the past stock/future prices and technique indicators, such as KD, RSI, and MACD, as features. Very few studies consider the forum messages or bargaining chips as stock and/or futures price prediction features. In this thesis, discussion messages from both PTT and CMoney forums are converted into daily sentimental vectors using the retrained BERT. The daily sentimental vector as well as three bargaining chips are then used as features to train the GRU and TCN models. The experiment results show that the TCN performs better than the GRU-based RNN model in terms of MAE, MAPE, RMSE, and accuracy. In addition, both of the bargaining chips and forum messages are verified to be useful in the futures price prediction. The market simulations based on the historical futures price show that a simple investment strategy using the TCN model using techniques, bargaining chips, and forum messages can earn more than 7 times of the investment in the period of one year.

關鍵字(中)

★ 期貨預測
★ 時間卷積網路
★ 情緒分析

關鍵字(英)

★ Futures Prediction
★ Temporal Convolutional Network
★ Sentiment Analysis

論文目次

1 Introduction 1
2 RelatedWork 3
2.1 Deep Learning with Numerical Features . . . . . . . . . . . . . . . . . . . 3
2.2 Deep Learning with Graph Features . . . . . . . . . . . . . . . . . . . . . . 4
2.3 Sentiment Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
3 Preliminary 7
3.1 Deep Learning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
3.1.1 Multi-Layer Perceptrons . . . . . . . . . . . . . . . . . . . . . . . . 7
3.1.2 Convolutional Neural Networks . . . . . . . . . . . . . . . . . . . . 8
3.1.3 Recurrent Neural Networks . . . . . . . . . . . . . . . . . . . . . . 9
3.1.4 Temporal Convolutional Networks . . . . . . . . . . . . . . . . . . . 10
3.2 Bidirectional Encoder Representations from Transformers . . . . . . . . . . 11
3.3 Techniques to Reduce Overfitting . . . . . . . . . . . . . . . . . . . . . . . 12
3.3.1 Early Stopping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
3.3.2 Dropout . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
3.3.3 L1 and L2 Regularizations . . . . . . . . . . . . . . . . . . . . . . . 15
3.4 Model Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
3.4.1 Confusion Metrics . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
3.4.2 Evaluation Metrics . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
3.5 Financial Indicators and Bargaining Chips . . . . . . . . . . . . . . . . . . 19
4 Design 22
4.1 Problem Formulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
4.2 System Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
4.3 Data Collection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
4.3.1 Historical Data of TAIEX Futures . . . . . . . . . . . . . . . . . . . 24
4.3.2 TAIEX Futures Forum . . . . . . . . . . . . . . . . . . . . . . . . . 25
4.4 Data Pre-processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
4.4.1 Retrained BERT Model . . . . . . . . . . . . . . . . . . . . . . . . 25
4.4.2 Evaluation Metrics of retrained BERT . . . . . . . . . . . . . . . . 26
4.4.3 Random Oversampling . . . . . . . . . . . . . . . . . . . . . . . . . 28
4.4.4 Daily Sentimental Vectors . . . . . . . . . . . . . . . . . . . . . . . 28
4.5 Model Construction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
5 Performance 31
5.1 Evaluation Metrics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
5.1.1 Mean Absolute Error . . . . . . . . . . . . . . . . . . . . . . . . . . 31
5.1.2 Root Mean Square Error . . . . . . . . . . . . . . . . . . . . . . . . 32
5.1.3 Mean Absolute Percentage Error . . . . . . . . . . . . . . . . . . . 32
5.1.4 Accuracy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
5.1.5 Market Simulation . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
5.2 Model Tuning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
5.2.1 Model Tuning for GRU Models . . . . . . . . . . . . . . . . . . . . 37
5.2.2 Model Tuning for TCN Models . . . . . . . . . . . . . . . . . . . . 40
5.3 Experimental Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
6 Conclusions and Future Works 45
Reference 46

參考文獻

[1] Xor. https://www.pcmag.com/encyclopedia/term/xor.
[2] Md Saiful Islam Arnab Sen Sharma, Maruf Ahmed Mridul. Automatic detection of satire in bangla documents: A cnn approach based on hybrid feature extraction model. 2019 International Conference on Bangla Speech and Language Processing (ICBSLP), pages 1 – 5, 2019.
[3] Shaojie Bai, J. Zico Kolter, and Vladlen Koltun. An empirical evaluation of generic convolutional and recurrent networks for sequence modeling. arXiv:1803.01271, 2018.
[4] Businessinsider.com. Sp500. https://markets.businessinsider.com/index/s&p_ 500.
[5] JAMES CHEN. Systematic investment plan. https://www.investopedia.com/ terms/s/systematicinvestmentplan.asp, January 2020.
[6] Junyoung Chung, Caglar Gulcehre, KyungHyun Cho, and Yoshua Bengio. Gru. https://arxiv.org/abs/1412.3555.
[7] CMoney Corporation. Cmoney. http://www.cmoney.com.tw/default.asp.
[8] CMoney Corporation. Cmoneyforum. https://www.cmoney.tw/follow/channel/ entry.
[9] Farhana Zulkernine Dev Shah, Haruna Isah. Predicting the effects of news sentiments on the stock market. 2018 IEEE International Conference on Big Data (Big Data), pages 4705–4708, 2018.
[10] AKHILESH GANTI. Openinterest. https://www.investopedia.com/terms/o/ openinterest.asp.
[11] Inc Google. Bert. https://github.com/google-research/bert.
[12] Inc Google. Wavenet. http://sergeiturukin.com/2017/03/02/wavenet.html.
[13] Jorge A. Castanon Hao Wang. Sentiment expression via emoticons on social media. 2015 IEEE International Conference on Big Data (Big Data), pages 2404–2408, 2015.
[14] Min-Te Sun Hu-Hsiang Yeh. Coal price prediction using financial indices. 2019 International Conference on Technologies and Applications of Artificial Intelligence (TAAI), pages 1–6, 2019.
[15] Investopedia.com. Nasdaq. https://www.investopedia.com/terms/n/ nasdaqcompositeindex.asp.
[16] Renuka Joshi. Interpretation of performance measures. https://blog.exsilio. com/all/accuracy-precision-recall-f1-score-interpretation-of-performance-measures/, Sep 2016.
[17] Fangyan Dai Kai Chen, Yi Zhou. A lstm-based method for stock returns prediction: A case study of china stock market. 2015 IEEE International Conference on Big Data (Big Data), pages 2823–2824, 2015.
[18] Mutasem khalil Alsmadi, Shahrul Azman Noah Khairuddin Bin Omar, and Ibrahim Almarashdah. Performance comparison of multi-layer perceptron (back propagation, delta rule and perceptron) algorithms in neural networks. 2009 IEEE International Advance Computing Conference DOI, pages 296 – 299, 2009.
[19] Sookasame Chalita Lounnapha Sayavong, Zhongdong Wu. Research on stock price prediction method based on convolutional neural network. 2019 International Conference on Virtual Reality and Intelligent Systems (ICVRIS), pages 173–176, 2019.
[20] A. Murat Ozbayoglu M. Ugur Gudelek, S. Arda Boluk. Artificial neural networks architectures for stock price prediction: Comparisons and applications. Circuits Systems and Signal Processing, 10:1 – 8, 2016.
[21] A. Murat Ozbayoglu M. Ugur Gudelek, S. Arda Boluk. A deep learning based stock trading model with 2-d cnn trend detection. 2017 IEEE Symposium Series on Computational Intelligence (SSCI), pages 1 – 8, 2017. [22] Ashutosh Pandey and DeLiang Wang. Tcnn: Temporal convolutional neural network for real-time speech enhancement in the time domain. 2019.
[23] Jason Brownlee PhD. Random oversampling. https://machinelearningmastery. com/random-oversampling-and-undersampling-for-imbalanced-classification/, January 2021.
[24] PTT. Ptt. https://www.ptt.cc/bbs/Stock/index.html.
[25] J. Schmidhuber S. Hochreiter. Long short-term memory. Neural Computation, 9(8):1735–1780, 1997.
[26] Gordon Scott. Futures exchange. https://www.investopedia.com/terms/f/ futuresexchange.asp.
[27] Tong Sun, Jia Wang, Pengfei Zhang, Yu Cao, Benyuan Liu, and Degang Wang. Predicting stock price returns using microblog sentiment for chinese stock market.2017 3rd International Conference on Big Data Computing and Communications (BIGCOM), pages 87–96, 2017.
[28] Hyung Soon Kim Sunchan Park, Yongwon Jeong. Multiresolution cnn for reverberant speech recognition. 2017 20th Conference of the Oriental Chapter of the International Coordinating Committee on Speech Databases and Speech I/O Systems and Assessment (O-COCOSDA), pages 1 – 4, 2017.
[29] TAIEX. Taiex futures. https://www.taifex.com.tw/enl/eng2/tX.
[30] Taifex. Taifex. https://www.taifex.com.tw/enl/eIndex.
[31] T.-T. Teoh, W. T. Lim, K. W. Koh, J. J. Soh, T. Tan, S.Y. Liu, and Y.-Y. Nguwi. From technical analysis to text analytics: Stock and index prediction with gru. 2019 IEEE International Conference on Cybernetics and Intelligent Systems (CIS) and IEEE Conference on Robotics, Automation and Mechatronics (RAM), pages 496 – 500, 2019.
[32] Yi Liu Tingwei Gao, Yueting Chai. Applying long short term momory neural networks for predicting stock closing price. Transportation Research Part C: Emerging Technologies, pages 575–578, 2017.
[33] TWSE. Twse. https://www.twse.com.tw/zh/.
[34] TWSE. Systematic investment plan. https://www.twse.com.tw/en/, October 1961.
[35] WIKI. Taiex. https://en.wikipedia.org/wiki/Taiwan_Capitalization_ Weighted_Stock_Index.
[36] Inc Yahoo. Bbl. https://finance.yahoo.com/quote/BBL.BK/
[37] Inc Yahoo. Cpall. https://finance.yahoo.com/quote/CPALL.BK/.
[38] Inc Yahoo. Ppt. https://finance.yahoo.com/quote/PTT.BK/.
[39] Baoyou Zheng Yiwei Liu, Zhiping Wang. Application of regularized gru-lstm model in stock price prediction. 2019 IEEE 5th International Conference on Computer and Communications (ICCC), pages 1886–1890, 2019.
[40] Fisher Yu and Vladlen Koltun. Multi-scale context aggregation by dilated convolutions. 2016.
[41] Kuangrong Hao Zhenhua Yang, Xin Cai, Lei Chen, and Lihong Ren. Prediction of stock trading signal based on multi-indicator channel convolutional neural networks. 2019 IEEE 8th Data Driven Control and Learning Systems Conference (DDCLS), pages 912–917, 2019.

指導教授

孫敏德(Min-Te Sun)

審核日期

2021-1-28

推文