進階搜尋


   電子論文尚未授權公開,紙本請查館藏目錄
(※如查詢不到或館藏狀況顯示「閉架不公開」,表示該本論文不在書庫,無法取用。)
系統識別號 U0026-0808201802094400
論文名稱(中文) 時序式神經網路的文本分析應用於比特幣價格預測
論文名稱(英文) Text Analysis for Prediction of Bitcoin Price by Sequence Neural Network Model
校院名稱 成功大學
系所名稱(中) 資訊工程學系
系所名稱(英) Institute of Computer Science and Information Engineering
學年度 106
學期 2
出版年 107
研究生(中文) 林宇德
研究生(英文) Yu-De Lin
學號 P76054680
學位類別 碩士
語文別 英文
論文頁數 54頁
口試委員 指導教授-鄭憲宗
口試委員-盧文祥
口試委員-李麗華
口試委員-陳嘉平
中文關鍵字 文本分析  市場預測  Sequence to Sequence 
英文關鍵字 Text Analysis  market prediction  Sequence to Sequence 
學科別分類
中文摘要 隨著人工智慧加速發展,許多人想利用AI來預測市場趨勢,而數位貨幣價格以比特幣、乙太幣為首又在去年(2017年)狂飆數十甚至數百倍,媒體、社交網站上討論熱度越來越熱,人們自然得希望能用AI的方式預測虛擬幣市場。
本研究以推特貼文為訓練資料並以向量化的方式來表示一天的推文資訊(日向量)。首先清理Twitter原始資料,經過日向量模型把每天的推文轉換成日向量,最後以日向量為基礎用Sequence to Sequence模型進行比特幣漲跌幅進行預測。整個系統分別在日向量模型及Sequence to Sequence模型各採用了1個attention模型。
實驗結果可以看出隨著日向量維度上升預測準確率會略為上升,而SequenceDecoder model 的attention model可以較明顯的提升預測準確度。最後,我們個別分析7天的預測結果,發現預測越前面的天數(下一天,下兩天)準確率越高,這也和我們的直覺相符合。
英文摘要 With the accelerated development of artificial intelligence, some people want to use it to predict market trends. Simultaneously, digital currency, headed by Bitcoin and Ethereum, caught people’s attention because of its soaring price in last year. The reputation of digital currency get higher and higher in social media and traditional media. People certainly hope to use AI to predict the digital currency market.
In this research, we use Twitter posts as training data and vectored method to represent the tweet information (day vector) per day. After cleaning Twitter raw data, we converted the tweets in the giving day as day vector and feed the day vector to sequence to Sequence model use to predict the change of Bitcoin price. The entire system uses attention model in day vector model and the sequence to sequence model, respectively.
The experiments show that the prediction accuracy rise slightly by increasing day vector dimension and the attention model of the SequenceDecoder model can significantly improve the accuracy. Finally, we analyzed the 7-day predicted results individually and found that the accuracy decrease when predicting latter day. This meet our understanding that it is harder to predict the latter day than the near day.
論文目次 摘要 I
Abstract II
ACKNOWLEDGEMENT III
LIST OF FIGURES V
LIST OF TABLES VI
Chapter 1. Introduction and Motivation 1
1.1 Digital currency 1
1.2 Natural Language Processing with Deep Learning 2
1.3 Motivation 3
1.4 Thesis Overview 5
Chapter 2. Background and Related work 6
2.1 Recurrent Neural Networks 6
2.2 Sequence-to-Sequence model and Attention mechanism 9
2.3 Word2Vec 11
2.4 Relation between twitter mood and stock market 13
2.5 News-driven prediction 14
2.6 Stock prediction by using deep learning 16
Chapter 3. System design and Approach 18
3.1 Problem Description 18
3.2 System Design 20
3.2.1 Crawling Twitter Raw data and Data cleaning 20
3.2.2 Day Vector Model 23
3.2.3 Sequence to Sequence Model 27
3.2.4 Overall system 33
Chapter 4. Experiment 35
4.1 Experimental Environment and Settings 35
4.2 Experimental Result 36
4.3 Analysis Experiment Result 38
Chapter 5. Conclusion and Future work 42
References 44
參考文獻 [1] "wikipedia," [Online] Available: https://en.wikipedia.org/wiki/Ethereum.
[2] WU, Yonghui, et al. " Google's neural machine translation system: Bridging the gap between human and machine translation." arXiv Preprint arXiv, 2016.
[3] J. BOLLEN, H. MAO and X. ZENG, "Twitter mood predicts the stock market." Journal of Computational Science, p. 1-8, 2011.
[4] DING, Xiao, et al. "Deep learning for event-driven stock prediction.," Ijcai, p. 2327-2333, 2015.
[5] HU, Ziniu, et al. "Listening to chaotic whispers: A deep learning framework for news-oriented stock trend prediction." in Proceedings of the Eleventh ACM International Conference on Web Search and Data Mining. , 2018.
[6] wikipedia, "Fundamental_analysis" [Online] Available: https://en.wikipedia.org/wiki/Fundamental_analysis.
[7] wikipedia, "Technical_analysis" [Online] Available: https://en.wikipedia.org/wiki/Technical_analysis.
[8] "Blockchain Capital Bitcoin Survey Fall 2017" [Online] Available: https://www.survey.blockchain.capital/.
[9] "Bitcoin falls $1,000 after South Korea promises crackdown on trading" [Online] Available: https://www.theguardian.com/technology/2017/dec/28/bitcoin-falls-south-korea-crackdown-trading.
[10] S. HOCHREITER and J. SCHMIDHUBER, "Long short-term memory." Neural Computation, p. 1735-1780, 1997.
[11] R. PASCANU, T. MIKOLOV and Y. BENGIO, "On the difficulty of training recurrent neural networks." International Conference on Machine Learning, p. 1310-1318, 2013.
[12] I. SUTSKEVER, O. VINYALS and Q. V. LE, "Sequence to sequence learning with neural networks." Advances in Neural Information Processing Systems, p. 3104-3112, 2014.
[13] A. Karpathy, "The Unreasonable Effectiveness of Recurrent Neural Networks" 2015. [Online] Available: http://karpathy.github.io/2015/05/21/rnn-effectiveness/.
[14] D. BAHDANAU, K. CHO and Y. BENGIO, "Neural machine translation by jointly learning to align and translate." arXiv Preprint arXiv, 2014.
[15] T. e. a. MIKOLOV, “Distributed representations of words and phrases and their compositionality.,” Advances in Neural Information Processing Systems, p. 3111-3119, 2013.
[16] "word2vec" Google, [Online] Available: https://code.google.com/archive/p/word2vec/.
[17] Q. LE and T. MIKOLOV "Distributed representations of sentences and documents." International Conference on Machine Learning, p. 1188-1196, 2014.
[18] "github" [Online] Available: https://github.com/taspinar/twitterscraper.
[19] JAQUES, Natasha, et al. "Tuning recurrent neural networks with reinforcement learning" 2017.
[20] XIA, Yingce, et al. "Deliberation networks: Sequence generation beyond one-pass decoding." Advances in Neural Information Processing Systems, p. 1784-1794, 2017.
論文全文使用權限
  • 同意授權校內瀏覽/列印電子全文服務,於2022-12-31起公開。
  • 同意授權校外瀏覽/列印電子全文服務,於2022-12-31起公開。


  • 如您有疑問,請聯絡圖書館
    聯絡電話:(06)2757575#65773
    聯絡E-mail:etds@email.ncku.edu.tw