本研究運用包括Boosting技術、卷積神經網絡（CNN）、迴圈神經網絡（RNN）等一系列先進的機器學習與深度學習技術，預測比特幣在未來一百分鐘內價格的漲跌趨勢。透過應用基於互信息的特徵篩選法，本研究顯著提高了對比特幣價格波動的預測精度。此外，本研究創新性地將可解釋神經網絡（ENN）引入至加密貨幣預測領域，該模型結合了高維度數據處理與傳統加法模型的可解釋性，有效平衡了預測的透明度與準確性。在訓練階段，ENN模型透過其子網絡調整特徵權重，以適應新數據，從而為加密貨幣市場預測提供決策上的重要支持。通過創建結合交易量與價格的新變量及調整分類門檻，本研究準確預測了價格上升的趨勢。本研究成果不僅豐富了金融科技的分析工具，也為未來研究及實踐應用開闢了新的道路。

This study utilizes a range of advanced machine learning and deep learning techniques, including Boosting technology, Convolutional Neural Networks (CNN), and Recurrent Neural Networks (RNN), to predict the trends of Bitcoin price movements within the next 100 minutes. By applying a feature selection method based on Mutual Information, this research significantly enhances the accuracy of predictions for Bitcoin price fluctuations. Additionally, this study innovatively introduces the Explainable Neural Network (ENN) into the domain of cryptocurrency forecasting. This model combines high-dimensional data processing capabilities with the interpretability of traditional additive models, effectively balancing transparency and accuracy in predictions. The ENN model adjusts feature weights through its subnetworks during the training process, thus adapting to new data and providing crucial decision support for the cryptocurrency market. This structured adaptation during training allows the model to effectively capture and respond to complex patterns in data, enhancing both accuracy and interpretability. By creating a new variable that integrates trading volume and price, along with adjusting the classification threshold, this research accurately predicts upward price trends. The outcomes not only enrich the analytical tools in the field of financial technology but also pave new avenues for future research and practical applications.

論文審定書 i
致 謝 ii
摘 要 iv
Abstract v
TABLE OF CONTENTS vi
LIST OF FIGURES viii
LIST OF TABLES ix
Chapter 1 Introduction 1
1.1 Background 1
1.2. Price Determinants of Cryptocurrency 3
1.3. Recent Advancements in Machine learning and Deep learning 5
1.4. Objective and Main Results 7
1.5. Major Contributions 9
1.6. Thesis Outline 10
Chapter 2 Literature Review 11
2.1. Machine Learning and Deep Learning in Asset Pricing 11
2.2. Machine Learning and Deep Learning in Cryptocurrency Pricing 12
Chapter 3 Methodology 16
3.1. Boosting 16
3.1.1. Introduction to XGBoost 16
3.1.2. Introduction to CatBoost 18
3.2. CNN & Transformer 20
3.2.1. One-Dimensional ConvolutionEmbedding Transformer(OCET) 20
3.2.2. Convolutional Neural Network (CNN) 21
3.2.3. Transformer 22
3.3. Recurrent Neural Networks 23
3.3.1. Long Short-Term Memory (LSTM) 23
3.3.2. Gated Recurrent Units (GRUs) 25
3.4. Explainable Neural Network (ENN) 27
3.4.1. Overview 27
3.4.2. Model Architecture 28
3.5. feature selection 29
Chapter 4 Data and Evaluation Metrics 31
4.1. Data 31
4.2. Model Input and Output Specification 32
4.2.1. Model Input and Output 32
4.2.2. New Variable Creation 35
4.3. Data processing 37
4.4. Metrics 37
Chapter 5 Emperical Results 40
5.1. Overview of Model Performance 40
5.2. Analysis of High-Confidence Predictions 43
Chapter 6 Robustness Check 46
6.1 Setting a 60-Minute Target Variable 46
6.2 Specialized Testing on Ether 47
Chapter 7 Conclusion and Future Research Suggestions 50
References 52
Appendix 1: Analysis of Variable Importance 56

Agarwal, R., Melnick, L., Frosst, N., Zhang, X., Lengerich, B., Caruana, R., & Hinton, G. E. (2021). Neural additive models: Interpretable machine learning with neural nets. Advances in Neural Information Processing Systems, 34, 4699-4711.
Akyildirim, E., Goncu, A., & Sensoy, A. (2021). Prediction of cryptocurrency returns using machine learning. Annals of Operations Research, 297, 3-36.
Bybee, L., Kelly, B. T., Manela, A., & Xiu, D. (2021). Business news and business cycles (No. w29344). National Bureau of Economic Research.
Cai, C. X., & Zhao, R. (2024). Salience theory and cryptocurrency returns. Journal of Banking & Finance, 159, 107052.
CatBoost: unbiased boosting with categorical features. Advances in Neural Information Processing Systems, 31.
Chen, L., Pelger, M., & Zhu, J. (2024). Deep learning in asset pricing. Management Science, 70(2), 714-750.
Chen, T., & Guestrin, C. (2016, August). Xgboost: A scalable tree boosting system. In Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (pp. 785-794).
Cho, K., Van Merriënboer, B., Gulcehre, C., Bahdanau, D., Bougares, F., Schwenk, H., & Bengio, Y. (2014). Learning phrase representations using RNN encoder-decoder for statistical machine translation. arXiv preprint arXiv:1406.1078.
D’Amato, V., Levantesi, S., & Piscopo, G. (2022). Deep learning in predicting cryptocurrency volatility. Physica A: Statistical Mechanics and its Applications, 596, 127158.
Dudek, G., Fiszeder, P., Kobus, P., & Orzeszko, W. (2024). Forecasting cryptocurrencies volatility using statistical and machine learning methods: A comparative study. Applied Soft Computing, 151, 111132.
Fang, F., Chung, W., Ventre, C., Basios, M., Kanthan, L., Li, L., & Wu, F. (2024). Ascertaining price formation in cryptocurrency markets with machine learning. The European Journal of Finance, 30(1), 78-100.
Fang, F., Ventre, C., Basios, M., Kanthan, L., Martinez-Rego, D., Wu, F., & Li, L. (2022). Cryptocurrency trading: a comprehensive survey. Financial Innovation, 8(1), 13.
Fu, Y., Zhou, M., & Zhang, L. (2024). DAM: A universal dual attention mechanism for multimodal timeseries cryptocurrency trend forecasting. arXiv preprint arXiv:2405.00522.
Gkillas, K., Tantoula, M., & Tzagarakis, M. (2021). Transaction activity and bitcoin realized volatility. Operations Research Letters, 49(5), 715-719.
Gu, S., Kelly, B., & Xiu, D. (2020). Empirical asset pricing via machine learning. The Review of Financial Studies, 33(5), 2223-2273.
Gu, S., Kelly, B., & Xiu, D. (2021). Autoencoder asset pricing models. Journal of Econometrics, 222(1), 429-450.
Guijarro-Ordonez, J., Pelger, M., & Zanotti, G. (2021). Deep learning statistical arbitrage. arXiv preprint arXiv:2106.04028.
Härdle, W. K., Harvey, C. R., & Reule, R. C. (2020). Understanding cryptocurrencies. Journal of Financial Econometrics, 18(2), 181-208.
Hochreiter, S., & Schmidhuber, J. (1997). Long short-term memory. Neural Computation, 9(8), 1735-1780.
Hsiao, H. T., Wang, C. W., Liu, I. C., & Kung, K. L. (2024). Mortality improvement neural-network models with autoregressive effects. The Geneva Papers on Risk and Insurance-Issues and Practice, 1-21.
Jiang, J., Kelly, B., & Xiu, D. (2023). (Re‐) Imag (in) ing price trends. The Journal of Finance, 78(6), 3193-3249.
Khedr, A. M., Arif, I., El‐Bannany, M., Alhashmi, S. M., & Sreedharan, M. (2021). Cryptocurrency price prediction using traditional statistical and machine‐learning techniques: A survey. Intelligent Systems in Accounting, Finance and Management, 28(1), 3-34.
Li, Z., Han, Z., Cai, J., Huang, Z., Huang, S., & Kong, L. (2023). Profiling Cryptocurrency Influencers With Few-shot Learning via Contrastive Learning. In CLEF (Working Notes) (pp. 2695-2701).
Liu, X. Y., Yang, H., Gao, J., & Wang, C. D. (2021, November). FinRL: Deep reinforcement learning framework to automate trading in quantitative finance. In Proceedings of the Second ACM International Conference on AI in Finance (pp. 1-9).
Liu, Y., Tsyvinski, A., & Wu, X. (2022). Common risk factors in cryptocurrency. The Journal of Finance, 77(2), 1133-1177.
Makarov, I., & Schoar, A. (2020). Trading and arbitrage in cryptocurrency markets. Journal of Financial Economics, 135(2), 293-319.
Ntakaris, A., Magris, M., Kanniainen, J., Gabbouj, M., & Iosifidis, A. (2018). Benchmark dataset for mid‐price forecasting of limit order book data with machine learning methods. Journal of Forecasting, 37(8), 852-866.
Oreshkin, B. N., Carpov, D., Chapados, N., & Bengio, Y. (2019). N-BEATS: Neural basis expansion analysis for interpretable time series forecasting. arXiv preprint arXiv:1905.10437.
Park, J., & Seo, Y. S. (2022). A deep learning-based action recommendation model for cryptocurrency profit maximization. Electronics, 11(9), 1466.
Prokhorenkova, L., Gusev, G., Vorobev, A., Dorogush, A. V., & Gulin, A. (2018). CatBoost: unbiased boosting with categorical features. Advances in Neural Information Processing Systems, 31.
Schapire, R. E. (2003). The boosting approach to machine learning: An overview. Nonlinear Estimation and Classification, 149-171.
Schnaubelt, M. (2022). Deep reinforcement learning for the optimal placement of cryptocurrency limit orders. European Journal of Operational Research, 296(3), 993-1006.
Vaughan, J., Sudjianto, A., Brahimi, E., Chen, J., & Nair, V. N. (2018). Explainable neural networks based on additive index models. arXiv preprint arXiv:1806.01933.
Wu, S., Irsoy, O., Lu, S., Dabravolski, V., Dredze, M., Gehrmann, S., ... & Mann, G. (2023). Bloomberggpt: A large language model for finance. arXiv preprint arXiv:2303.17564.
Xu, W., Liu, W., Wang, L., Xia, Y., Bian, J., Yin, J., & Liu, T. Y. (2021). Hist: A graph-based framework for stock trend forecasting via mining concept-oriented shared information. arXiv preprint arXiv:2110.13716.
Yang, P., Fu, L., Zhang, J., & Li, G. (2022, December). OCET: One-Dimensional Convolution Embedding Transformer for Stock Trend Prediction. In International Conference on Bio-Inspired Computing: Theories and Applications (pp. 370-384). Singapore: Springer Nature Singapore.