Loading...
|
Please use this identifier to cite or link to this item:
https://nccur.lib.nccu.edu.tw/handle/140.119/130974
|
| Title: | 具解釋性機器學習模型：加密貨幣價格預測與多面向影響因子
Interpretable Machine Learning for Cryptocurrencies Price Prediction with Multidimensional Factors |
| Authors: | 許聖謙
Hsu, Sheng-Chien |
| Contributors: | 莊皓鈞
周彥君
許聖謙
Hsu, Sheng-Chien |
| Keywords: | 加密貨幣
價格預測
白箱模型
Cryptocurrencies
Price Predictions
White-box model |
| Date: | 2020 |
| Issue Date: | 2020-08-03 17:35:00 (UTC+8) |
| Abstract: | 加密貨幣自中本聰發表比特幣與區塊鏈相關應用後逐漸開始蓬勃發展，許多加密貨幣開始透過Initial Coin Offering (ICO)方式發行，提供換取產品與服務、獲得監管權利以及作為交易媒介的功能。在金融市場中，加密貨幣被視為金融商品，其價格具有高波動與高報酬的特性，且相較於股票，加密貨幣沒有財務報表資訊，價格較容易受到外部因素影響，要如何預測加密貨幣價格與了解波動因素對投資人來說成為重要課題。
本研究使用統計機器學習與多面向因子建構具解釋性的白箱模型，與無法直接得知模型運作的黑箱模型相比，白箱模型有助於理解模型輸入和輸出之間關係。本研究使用兼具解釋能力與良好預測能力的Lasso Regression來建構白箱模型，並納入四種主要變數，分別是代表市場影響力的經濟指標、呈現貨幣之間相互影響的加密貨幣價格、反映大眾未來期待的搜尋引擎指標與新聞情緒指標。接著與黑箱模型Random Forest、XGBoost、Deep Neural Network以及時間序列ARIMA分析進行結果比較，發現白箱模型能夠達到其他模型的預測準確度。除此之外，本研究也以高維度Vector Autoregression系統化地分析變數之間的關係，並使用視覺化方法解釋影響加密貨幣價格的重要因素。本研究主要貢獻包含探討白箱模型在價格預測上的適用性與了解價格影響因素，提供未來相關研究與投資決策的參考依據。
Satoshi Nakamoto published bitcoin and blockchain in 2009, since then, cryptocurrencies have gradually become more and more popular. The main functions of cryptocurrencies are providing products and services in exchange, obtaining regulatory rights, and functioning as a trading medium. Furthermore, cryptocurrencies are regarded as financial commodities with high volatility and high returns in the financial market. Compared with stocks, there is no financial statement information for cryptocurrencies. Therefore, the prices of cryptocurrencies might be more susceptible to external factors. Predicting the price of cryptocurrencies becomes an important issue for investors and the main goal of this study.
This study uses two white-box models which help to understand the relationship between model inputs and outputs as main methods. The first one is the Lasso Regression. As a white-box model, it has both explanatory power and good predictive power. The second method is the high-dimensional vector autoregression. It systematically analyzes the relationship between a large number of variables. We include four main variables, which are economic indicators, cryptocurrency prices, search engine indicators, and news sentiment indicators in the models. After model construction, we compare the accuracy with the black box models - Random Forest, XGBoost, Deep Neural Network, and time series ARIMA analysis. We find out that the white-box models reach the prediction accuracy of other complicated models. Furthermore, we use visualization methods to explain the important factors that affect the price of cryptocurrencies.
The main contributions of this research include exploring the applicability of the white-box model in price prediction, understanding the price influencing factors, and providing a reference for future related research and investment decisions. |
| Reference: | Abraham, J., Higdon, D., & Nelson, J. (2018). Cryptocurrency price prediction using tweet volumes and sentiment analysis. SMU Data Science Review, 1(3).
Adadi, A., & Berrada, M. (2018). Peeking Inside the Black-Box: A Survey on Explainable Artificial Intelligence (XAI). IEEE Access, 6, 52138–52160.
Alessandretti, L., Elbahrawy, A., Aiello, L. M., & Baronchelli, A. (2018). Anticipating Cryptocurrency Prices Using Machine Learning. Complexity, 2018.
Barnwal, A., Bharti, H. P., Ali, A., & Singh, V. (2019). Stacking with Neural Network for Cryptocurrency investment. 2019 New York Scientific Data Summit, NYSDS (NYSDS), 1–5.
Begušić, S., Kostanjčar, Z., Eugene Stanley, H., & Podobnik, B. (2018). Scaling properties of extreme price fluctuations in Bitcoin markets. Physica A: Statistical Mechanics and Its Applications, 510, 400–406.
Breiman, L. (2001). Random forests. Machine Learning, 45(1), 5–32.
Cai, Z., Lim, E. T. K., Liu, F., Tan, C. W., & Zheng, Z. (2018). Unraveling the effects of google search on volatility of cryptocurrencies. In International Conference on Information Systems 2018, ICIS 2018.
Chen, T., & Guestrin, C. (2016). XGBoost: A scalable tree boosting system. Proceedings of the 22nd Acm Sigkdd International Conference on Knowledge Discovery and Data Mining, 785–794.
Chen, Z., Li, C., & Sun, W. (2020). Bitcoin price prediction using machine learning: An approach to sample dimension engineering. Journal of Computational and Applied Mathematics, 365, 112395.
Ciaian, P., Rajcaniova, M., & Kancs, d’Artis. (2016). The economics of BitCoin price formation. Applied Economics, 48(19), 1799–1815.
Conrad, C., Custovic, A., & Ghysels, E. (2018). Long- and Short-Term Cryptocurrency Volatility Components: A GARCH-MIDAS Analysis. Journal of Risk and Financial Management, 11(2), 23.
Doran, D., Schulz, S., & Besold, T. R. (2018). What does explainable AI really mean? A new conceptualization of perspectives. CEUR Workshop Proceedings, 2071.
Dyhrberg, A. H. (2016). Bitcoin, gold and the dollar - A GARCH volatility analysis. Finance Research Letters, 16, 85–92.
Guesmi, K., Saadi, S., Abid, I., & Ftiti, Z. (2019). Portfolio diversification with virtual currency: Evidence from bitcoin. International Review of Financial Analysis, 63, 431–437.
He, D., Habermeier, K., Leckow, R., Haksar, V., Almeida, Y., Kashima, M., Kyriakos-Saad, N., Oura, H., Saadi Sedik, T., Stetsenko, N., Verdugo-Yepes, C., Viñals, J., Tiwari, S., Perry, V., Diaz-Kalan, F., Iorgova, S., Pampolina, J., Rendak, N., Strandquist, A., … Jagatsing, K. (2016). Virtual Currencies and Beyond: Initial Considerations INTERNATIONAL MONETARY FUND Monetary and Capital Markets, Legal, and Strategy and Policy Review Departments Virtual Currencies and Beyond: Initial Considerations. Staff Discussion Notes No. 16/3.
Hileman, G., & Rauchs, M. (2017). 2017 Global Cryptocurrency Benchmarking Study. Cambridge Centre for Alternative Finance, 33.
Loughran, T., & Mcdonald, B. (2011). When Is a Liability Not a Liability? Textual Analysis, Dictionaries, and 10-Ks. Journal of Finance, 66(1), 35–65.
McNally, S., Roche, J., & Caton, S. (2018). Predicting the Price of Bitcoin Using Machine Learning. Proceedings - 26th Euromicro International Conference on Parallel, Distributed, and Network-Based Processing, PDP 2018, 339–343.
Nicholson, W. B., Wilms, I., Bien, J., & Matteson, D. S. (2014). High Dimensional Forecasting via Interpretable Vector Autoregression. ArXiv Preprint ArXiv: 1412.5250.
Ribeiro, M. T., Singh, S., & Guestrin, C. (2016). “Why should i trust you?” Explaining the predictions of any classifier. Proceedings of the ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 13-17-Augu, 1135–1144.
Robnik-Šikonja, M., & Kononenko, I. (2008). Explaining classifications for individual instances. IEEE Transactions on Knowledge and Data Engineering, 20(5), 589–600.
Rognone, L., Hyde, S., & Zhang, S. (2020). News Sentiment in the Cryptocurrency Market: an Empirical Comparison with Forex. International Review of Financial Analysis, 69, 101462.
Samek, W., Wiegand, T., & Müller, K.-R. (2017). Explainable Artificial Intelligence: Understanding, Visualizing and Interpreting Deep Learning Models. ArXiv, 1708.08296.
Selmi, R., Mensi, W., Hammoudeh, S., & Bouoiyour, J. (2018). Is Bitcoin a hedge, a safe haven or a diversifier for oil price movements? A comparison with gold. Energy Economics, 74, 787–801.
Sims, C. A. (1980). Macroeconomics and Reality. Econometrica, 48(1), 1.
Smeekes, S., & Wijler, E. (2018). Macroeconomic forecasting using penalized regression methods. International Journal of Forecasting, 34(3), 408–430.
Song, S., & Bickel, P. J. (2011). Large Vector Auto Regressions. 1–28.
Sovbetov, Y. (2018). Factors Influencing Cryptocurrency Prices: Evidence from Bitcoin, Ethereum, Dash, Litcoin, and Monero. Journal of Economics and Financial Analysis, 2(2), 1–27.
Sun, X., Liu, M., & Sima, Z. (2019). A novel cryptocurrency price trend forecasting model based on LightGBM. Finance Research Letters, 32, 101084.
Tibshirani, R. (1996). Regression Shrinkage and Selection Via the Lasso. Journal of the Royal Statistical Society: Series B (Methodological), 58(1), 267–288.
Velankar, S., Valecha, S., & Maji, S. (2018). Bitcoin price prediction using machine learning. International Conference on Advanced Communication Technology, ICACT, 2018-February, 144–147.
Zeng, J., Ustun, B., & Rudin, C. (2017). Interpretable classification models for recidivism prediction. Journal of the Royal Statistical Society. Series A: Statistics in Society, 180(3), 689–722. |
| Description: | 碩士
國立政治大學
資訊管理學系
107356001 |
| Source URI: | http://thesis.lib.nccu.edu.tw/record/#G0107356001 |
| Data Type: | thesis |
| DOI: | 10.6814/NCCU202001026 |
| Appears in Collections: | [資訊管理學系] 學位論文
|
Files in This Item:
| File |
Description |
Size | Format | |
|---|
| 600101.pdf | | 4061Kb | Adobe PDF2 | 0 | View/Open |
|
All items in 政大典藏 are protected by copyright, with all rights reserved.
|
