Bitcoin price prediction using various machine learning techniques

Bitcoin, the pioneering cryptocurrency, has garnered immense attention for its volatile price movements and disruptive potential. Its price is influenced by various factors, including market demand, investor sentiment, regulatory developments, and technological advancements. Predicting Bitcoin's price is challenging due to its decentralized nature and susceptibility to market speculation. However, analysts often utilize technical analysis, fundamental indicators, and macroeconomic trends to forecast potential price movements. While some anticipate continued growth driven by increasing institutional adoption and limited supply, others warn of potential corrections amid regulatory uncertainties and market fluctuations. In essence, Bitcoin's price prediction remains a subject of debate and speculation within the financial community.

1. Short term Bitcoin Price Prediction via machine learning Under the supervision of Ms.Supriya Panigrahy Assistant Professor, Dept Of CSE, C.V. Raman Global University Submitted By : Name : Rishabh Singh Reg No : 2101020771 Roll No : CSI21042 Group : 08 Semester : 6

2. OUTLINES INTRODUCTION LITERATURE REVIEW PATENT SEARCH TECHNICAL DETAILS RESULT CONCLUSION REFERENCE

3. Bitcoin is a decentralized digital currency introduced in 2008 by Satoshi Nakamoto. It operates on a peer-to-peer network without the need for intermediaries, utilizing blockchain technology for secure transactions. With a finite supply capped at 21 million coins, it's often referred to as digital gold and has gained significant mainstream recognition. INTRODUCTION MOTIVATION Accurate Bitcoin price prediction entails forecasting future price movements through comprehensive analysis of historical data, market trends, sentiment, and fundamental indicators. This study investigates the use of machine learning, a productive technique in other domains, to the prediction of Bitcoin prices with the goal of improving financial forecasting.

4. ADVATAGES Processing diverse data Reduced human bias Improved Accuracy APPLICATION Market Sentiment Analysis Algorithmic Trading CONCLUSION The study analyzes short-term Bitcoin market predictability using diverse machine learning models across different timeframes. Models yield statistically significant predictions, improving with longer forecasts.

5. Publisher name, Journal year, Volm. Issue, Page Title of the paper & Author’s name Data Source Used Techniques used Contribution Conclusion of future scope Elsevier B.V, Journal of Finance and Data Science, pages 21, (2021) Short-term bitcoin market prediction via machine learning . Patrick Jaquart, David Dann,Christof Weinhardt Bloomberg, Twitter, and Blockchain.com Deep Learning Models - Neural Networks. Random Classifier Gradient Boosting Classifiers Helps in analysing complex data efficiently, fast decision making and improved accuracy. predicting market value for long term and improving accuracy Junwei Chen, Journal of Risk and Financial Management (JRFM), pages 51, (2023) Analysis of Bitcoin Price Prediction Using Machine Learning. Junwei Chen Bitcoin price dataset Time series data Deep Learning Models - LSTM Random Forest Dense and Dropout Layers Achieve improved accuracy in identifying bitcoin prices. shortening the time interval of samples automation LITRATURE REVIEW

6. LITRATURE REVIEW Publisher name, Journal year, Volm. Issue, Page Title of the paper & Author’s name Data Source Used Techniques used Contribution Conclusion of future scope Wei Chen, Huilin Xu, Lifen Jia, Ying Gao 0169-2070/© 2020. Published by Elsevier B.V Machine learning model for Bitcoin exchange rate prediction using economic and technology determinants. Wei Chen, Huilin Xu, Lifen Jia, Ying Gao APIs Yahoo YQL Finance API Bitinfocharts.com ANN Random Forest LSTM The study contributes to the understanding of the predictive capacity of economic and technology determinants in forecasting the Bitcoin exchange rate. Future studies could explore the predictive capacity of economic and technology determinants for other cryptocurrencies J ournal of Computational and Applied Mathematics Year: 2020, Volume: 365 Bitcoin Price Prediction Using Machine Learning: An Approach to Sample Dimension Engineering Zheshi Chen, Chunhong Li, Wenjun Sun daily price data from CoinMarketCap.com Real-time 5-minute interval : Binance cryptocurrency exchange. Statistical Methods: Logistic Regression Linear Discriminant Analysis (LDA) Machine Learning Models: Long Short-Term Memory (LSTM) Random Forest (RF) XGBoost (XGB) Utilize high-dimensional features and simpler models for daily and high-frequency Bitcoin price prediction. Extending studies to various granularities and features is crucial for improving machine learning applications in cryptocurrency  markets.

7. Your logo PATENT SEARCH

8. TECHNICAL DETAILS RNN : Recurrent Neural Networks (RNNs) are a type of artificial neural network designed for sequential data processing. Unlike traditional feedforward neural networks, RNNs have connections that loop back on themselves, allowing them to maintain information across inputs. This looping mechanism enables RNNs to capture temporal dependencies in data, making them suitable for tasks like time series prediction, natural language processing, and speech recognition. Structure of RNN in a folded state. Structure of RNN in an unfolded state.

9. Long-short term memory: Long Short-Term Memory (LSTM) belongs to the category of gated recurrent neural networks (RNNs) architecture designed to address the vanishing gradient problem and specific focus on long-term memorization of information in sequential data. Widely used in tasks like neural language processing, speech recognition, and financial data analysis. LSTM memory block with input, forget, and output gate.

10. Gradient boosting classifiers (GBC): leverage multiple decision trees and train individual weights for each model. This method emphasizes better-adapted models in the final classification decision. GBCs, like random forests, aim to improve prediction accuracy by combining the outputs of multiple trees. RESULTS

11. CONCLUSION Using several machine learning models on four distinct prediction horizons, we empirically analyse the short-term predictability of the bitcoin market. All of the evaluated models, it turns out, produce statistically meaningful predictions. The predictive performance of the models increases with longer forecast horizons, with accuracies ranging from 50.9% to 56.0% in the binary market movement prediction. It is shown that gradient boosting classifiers and recurrent neural networks in particular function well for this kind of prediction. Technical aspects continue to be prevalently relevant for the majority of approaches when feature groups consisting of technical, blockchain-based, sentiment/interest-based, and asset-based features are compared. Longer prediction horizons seem to divide the relative relevance among several factors (transactions per second, weighted sentiment), with less recent technical features becoming more significant.

12. REFERENCES 1. Nakamoto S. Bitcoin: a peer-to-peer electronic cash system. Work Pap; 2008. https://bitcoin.org/bitcoin.pdf. 2. Beck R, Avital M, Rossi M, Thatcher JB. Blockchain technology in business and information systems research. Bus Inf Syst Eng. 2017;URL: 59:381e384. https://doi.org/10.1007/s12599-017-0505-1. 3. Gu S, Kelly B, Xiu D. Empirical asset pricing via machine learning. Rev Financ Stud. 2020; URL:33:2223e2273.https://doi.org/10.1093/rfs/hhaa009. 4. Feng G, Giglio S, Xiu D. Taming the factor zoo: a test of new factors. J Finance. 2020; URL:75:1327e1370.https://doi.org/10.1111/jofi.12883. 5. Jaquart P, Dann D, Martin C. Machine learning for bitcoin pricing - a structured literature review. In: WI 2020 Proceedings, GITO Verlag.2020; URL: 174e188. https://doi.org/10.30844/wi_2020_b4-jaquart. 6. Coinmarketcap; 2020. URL: https://coinmarketcap.com/. Accessed September 15, 2020. 7. Fama EF. Efficient capital markets: a review of theory and empirical work. J Finance. 1970; URL:25:383e417.https://doi.org/10.1111/j.1540-6261.1970.tb00518.x. 8. Lo AW. The adaptive markets hypothesis. In: Adaptive Markets. vol. 30. Princeton University Press; 2019; URL:176e221.https://doi.org/10.2307/j.ctvc7778k.9. 9. Fama EF, MacBeth JD. Risk, return, and equilibrium: empirical tests. J Polit Econ. 1973; URL:81:607e636.https://doi.org/10.1086/260061. 10. Fama EF, French KR. Dissecting anomalies. J Finance. 2007; URL: 63:1653e1678. https://doi.org/10.2139/ssrn.911960. 11. Fischer T, Krauss C. Deep learning with long short-term memory networks for financial market predictions. Eur J Oper Res. 2018;270:654e669. https://doi.org/10.1016/j.ejor.2017.11.054. 12. Krollner B, Vanstone B, Finnie G. Financial time series forecasting with machine learning techniques: a survey. In: Proceedings of European Symposium on Artificial Neural Networks: Computational and Machine Learning. Springer; 2010:1e7. 13. Fama EF. Market efficiency, long-term returns, and behavioral finance. J Financ Econ. 1997;49:283e306. https://doi.org/10.2139/ssrn.15108. 14. Grossman SJ, Stiglitz JE. On the impossibility of informationally efficient markets. Am Econ Rev. 1980;70:393e408. 15. Green J, Hand JRM, Zhang F. The supraview of return predictive signals. Rev Account Stud. 2012;18:692e730. https://doi.org/10.2139/ ssrn.2062464.

13. 16. Shleifer A, Vishny R. The limits of arbitrage. J Finance. 1995;52:35e55. https://doi.org/10.1111/j.1540-6261.1997.tb03807.x. 17. Schwert GW. Anomalies and market efficiency. In: Handbook of the Economics of Finance, Handbook of the Economics of Finance. Elsevier; 2003:939e974. https://doi.org/10.1016/S1574-0102(03)01024-0. 18. Glaser F, Zimmermann K, Haferkorn M, Weber MC, Siering M. Bitcoin-asset or currency? revealing users' hidden intentions. In: Proceedings of 22nd European Conference on Information Systems. 2014:1e15. 19. Dyhrberg AH. Bitcoin, gold and the dollar - a GARCH volatility analysis. Finance Res Lett. 2016;16:85e92. https://doi.org/10.1016/ j.frl.2015.10.008. 20. Burniske C, White A. Bitcoin: Ringing the Bell for a New Asset Class, Working Paper. 2017. 21. Hu AS, Parlour CA, Rajan U. Cryptocurrencies: stylized facts on a new investible instrument. Financ Manag. 2019;48:1049e1068. https:// doi.org/10.1111/fima.12300. 22. Urquhart A. The inefficiency of bitcoin. Econ Lett. 2016;148:80e82. https://doi.org/10.1016/j.econlet.2016.09.019. 23. Nadarajah S, Chu J. On the inefficiency of Bitcoin. Econ Lett. 2017;150:6e9. https://doi.org/10.1016/j.econlet.2016.10.033. 24. Bariviera AF. The inefficiency of Bitcoin revisited: a dynamic approach. Econ Lett. 2017;161:1e4. https://doi.org/10.1016/ j.econlet.2017.09.013. 25. Hurst HE. Long-term storage capacity of reservoirs. Trans Am Soc Civ Eng. 1951;116:770e799. 26. Vidal-Tom_x0001_as D, Iba~nez A. Semi-strong efficiency of bitcoin. Finance Res Lett. 2018;27:259e265. https://doi.org/10.1016/j.frl.2018.03.013. 27. Khuntia S, Pattanayak J. Adaptive market hypothesis and evolving predictability of bitcoin. Econ Lett. 2018;167:26e28. https://doi.org/ 10.1016/j.econlet.2018.03.005. 28. Karakoyun ES, Cibikdiken AO. Comparison of ARIMA time series model and LSTM deep learning algorithm for bitcoin price forecasting. In: Proceedings of 13th Multidisciplinary Academic Conference. 2018:171e180. 29. McNally S, Roche J, Caton S. Predicting the price of bitcoin using machine learning. In: Proceedings of 2018 Euromicro International Conference on Parallel, Distributed, and Network-Based Processing. 2018:339e343. https://doi.org/10.1109/PDP2018.2018.00060. 30. Huang J-Z, Huang W, Ni J. Predicting bitcoin returns using high-dimensional technical indicators. J Finance Data Sci. 2019;5:140e155. https://doi.org/10.1016/j.jfds.2018.10.001. 31. Chen Z, Li C, Sun W. Bitcoin price prediction using machine learning: an approach to sample dimension engineering. J Comput Appl Math. 2020;365, 112395. https://doi.org/10.1016/j.cam.2019.112395.

14. 32. Kubat M, Matwin S. Others, Addressing the curse of imbalanced training sets: one-sided selection. In: Proceedings of the 14th International Conference on MachineLearning. Citeseer, Morgan Kaufmann; 1997:179e186. 33. Barandela R, Valdovinos RM, S_x0001_anchez JS, Ferri FJ. The imbalanced training sample problem: under or over sampling?. In: Lecture Notes in Computer Science. Springer, Springer; 2004:806e814. https://doi.org/10.1007/978-3-540-27868-9_88. 34. Mudassir M, Bennbaia S, Unal D, Hammoudeh M. Time-series forecasting of Bitcoin prices using high-dimensional features: a machine learning approach. Neural Comput Appl. 2020:1e15. https://doi.org/10.1007/s00521-020-05129-6. 35. Dutta A, Kumar S, Basu M. A gated recurrent unit approach to bitcoin price prediction. J Risk Financ Manag. 2020;13. https://doi.org/ 10.3390/jrfm13020023. Article 23. 36. Jaquart P, Dann D, Weinhardt C. Using machine learning to predict short-term movements of the bitcoin market. In: Proceedings of 10th International Workshop on Enterprise Applications, Markets and Services in the Finance Industry. Springer; 2020:21e40. https://doi.org/ 10.1007/978-3-030-64466-6_2. 37. Poyser O. Exploring the dynamics of Bitcoin's price: a Bayesian structural time series approach. Eurasian Econ Rev. 2019;9:29e60. https:// doi.org/10.1007/s40822-018-0108-2. 38. Madan I, Saluja S, Zhao A. Automated Bitcoin Trading via Machine Learning Algorithms. Working Paper; 2015. 39. Smuts N. What drives cryptocurrency prices? An investigation of Google trends and telegram sentiment. ACM SIGMETRICS Perform Eval Rev. 2019;46:131e134. https://doi.org/10.1145/3308897.3308955. 40. McKinney W. Data structures for statistical computing in Python. In: Proceedings of 9th Python in Science Conference. vol. 445. TX: Austin; 2010:56e61. https://doi.org/10.25080/Majora-92bf1922-00a. 41. van der Walt S, Colbert SC, Varoquaux G. The NumPy array: a structure for efficient numerical computation. Comput Sci Eng. 2011;13:22e30. https://doi.org/10.1109/MCSE.2011.37. 42. Bird S, Klein E, Loper E. Natural Language Processing with Python: Analyzing Text with the Natural Language Toolkit. O'Reilly Media, Inc.; 2009. 43. Google, Google. Cloud Natural Language API; 2020. URL: https://cloud.google.com/natural-language/docs. 44. Chollet F. Keras; 2015. URL: https://keras.io/. 45. Abadi M, Barham P, Chen J, et al. Tensorflow: a system for large-scale machine learning. In: Proceedings of 12th USENIX Symposium on Operating Systems Design and Implementation. 2016:265e283. 46. Chen T, Guestrin C, Boost XG. In: Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. ACM; 2016:785e794. https://doi.org/10.1145/2939672.2939785. 47. Pedregosa F, Varoquaux G, Gramfort A, et al. Others, scikit-learn: machine learning in Python. J Mach Learn Res. 2011;12:2825e2830.

15. 48. Kumar A, Garg G. Sentiment analysis of multimodal twitter data. Multimed Tool Appl. 2019;78:24103e24119. https://doi.org/10.1007/ s11042-019-7390-1. 49. Symeonidis S, Effrosynidis D, Arampatzis A. A comparative evaluation of pre-processing techniques and their interactions for twitter sentiment analysis. Expert Syst Appl. 2018;110:298e310. https://doi.org/10.1016/j.eswa.2018.06.022. 50. Goodfellow I, Bengio Y, Courville A. Deep Learning. MIT press; 2016. 51. Takeuchi L, Lee Y-YA. Applying Deep Learning to Enhance Momentum Trading Strategies in Stocks. Working Paper; 2013. 52. Krauss C, Do XA, Huck N. Deep neural networks, gradient-boosted trees, random forests: statistical arbitrage on the S&P 500. Eur J Oper Res. 2017;259:689e702. https://doi.org/10.1016/j.ejor.2016.10.031. 53. Finnoff W, Hergert F, Zimmermann HG. Improving model selection by nonconvergent methods. Neural Networks. 1993;6:771e783. https:// doi.org/10.1016/S0893-6080(05)80122-4. 54. Prechelt L. Early stopping - but when?. In: Neural Networks: Tricks of the Trade. Springer; 2012:53e67. https://doi.org/10.1007/978-3-642- 35289-8_5. 55. Kingma DP, Ba J, Adam. A method for stochastic optimization. In: Proceedings of 3rd International Conference on Learning Representa_x0002_tions. 2015. 56. Srivastava N, Hinton G, Krizhevsky A, Sutskever I, Salakhutdinov R. Dropout: a simple way to prevent neural networks from overfitting. J Mach Learn Res. 2014;15:1929e1958. 57. Hornik K. Approximation capabilities of multilayer feedforward networks. Neural Networks. 1991;4:251e257. https://doi.org/10.1016/0893- 6080(91)90009-T. 58. Hammer B. On the approximation capability of recurrent neural networks. Neurocomputing. 2000;31:107e123. https://doi.org/10.1016/ S0925-2312(99)00174-5. 59. Hochreiter S, Schmidhuber J. Long short-term memory. Neural Comput. 1997;9:1735e1780. https://doi.org/10.1162/neco.1997.9.8.1735. 60. Graves A, Schmidhuber J. Framewise phoneme classification with bidirectional LSTM and other neural network architectures. Neural Networks. 2005;18:602e610. https://doi.org/10.1016/j.neunet.2005.06.042. 61. Graves A, Fern_x0001_andez S, Gomez F, Schmidhuber J. Connectionist temporal classification. In: Proceedings of the 23rd International Conference on Machine Learning. ACM Press; 2006:369e376. https://doi.org/10.1145/1143844.1143891. 62. Graves A, Mohamed A-r, Hinton G. Speech recognition with deep recurrent neural networks. In: 2013 IEEE International Conference on Acoustics, Speech and Signal Processing. IEEE; 2013:6645e6649. https://doi.org/10.1109/ICASSP.2013.6638947. 63. Graves A, Jaitly N. Towards end-to-end speech recognition with recurrent neural networks. In: Proceedings of International Conference on Machine Learning. 2014:1764e1772. 64. Graves A. Generating Sequences with Recurrent Neural Networks. Working Paper; 2013.

16. 65. Liwicki M, Graves A, Fern_x0003_andez S, Bunke H, Schmidhuber J. A novel approach to on-line handwriting recognition based on bidirectional long short-term memory networks. In: Proceedings of 9th International Conference on Document Analysis and Recognition. ICDAR; 2007. 66. Graves A, Liwicki M, Bunke H, Schmidhuber J, Fern_x0001_andez S. Unconstrained on-line handwriting recognition with recurrent neural networks. In: Proceedings of Advances in Neural Information Processing Systems Conference. 2008:577e584. 67. Graves A, Liwicki M, Fernandez S, Bertolami R, Bunke H, Schmidhuber J. A novel connectionist system for unconstrained handwriting recognition. IEEE Trans Pattern Anal Mach Intell. 2009;31:855e868. https://doi.org/10.1109/TPAMI.2008.137. 68. Eck D, Schmidhuber J. Finding temporal structure in music: blues improvisation with LSTM recurrent networks. In: Proceedings of the 12th IEEE Workshop on Neural Networks for Signal Processing. IEEE; 2002:747e756. https://doi.org/10.1109/NNSP.2002.1030094. 69. M€akinen M, Iosifidis A, Gabbouj M, Kanniainen J. Predicting Jump Arrivals in Stock Prices Using Neural Networks with Limit Order Book Data. 2018. https://doi.org/10.2139/ssrn.3165408. Working Paper. 70. Gers FA, Schraudolph NN, Schmidhuber J. Learning precise timing with LSTM recurrent networks. J Mach Learn Res. 2002;3:115e143. 71. G ers F. Learning to forget: continual prediction with LSTM. In: Proceedings of 9th International Conference on Artificial Neural. vol. 1999. IEEE; 1999:850e855. https://doi.org/10.1049/cp:19991218. 72. Chung J, Gu¨lçehre Ç, Cho K, Bengio Y. Empirical Evaluation of Gated Recurrent Neural Networks on Sequence Modeling. Working Paper; 2014. arXiv:1412.3555. 73. Olah C. Understanding lstm networks. URL: https://colah.github.io/posts/2015-08-Understanding-LSTMs/; 2015. Accessed September 1,2020. 74. Ho T. Random decision forests. In: Proceedings of 3rd International Conference on Document Analysis and Recognition. vol. 1. IEEE Comput. Soc. Press; 1995:278e282. https://doi.org/10.1109/ICDAR.1995.598994. 75. Friedman JH. Greedy function approximation: a gradient boosting machine. Ann Stat. 2001;5:1189e1232. 76. Diebold F, Mariano R. Comparing predictive accuracy. J Bus Econ Stat. 1994;20:134e144. https://doi.org/10.1198/073500102753410444. 77. Breiman L. Random forests. Mach Learn. 2001;45:5e32. https://doi.org/10.1023/A:1010933404324. 78. Fischer T, Krauss C, Deinert A. Statistical arbitrage in cryptocurrency markets. J Risk Financ Manag. 2019;12:31. https://doi.org/10.3390/ jrfm12010031. 79. Biais B, Bisiere C, Bouvard M, Casamatta C, Menkveld AJ. Equilibrium Bitcoin Pricing, Working Paper. https://doi.org/10.2139/ssrn.3261063; 2018. 80. Detzel A, Liu H, Strauss J, Zhou G, Zhu Y. Learning and Predictability via Technical Analysis: Evidence from Bitcoin and Stocks with Hard_x0002_To-Value Fundamentals. Financial Management In Press; 2020. https://doi.org/10.1111/fima.12310. 81. Jegadeesh N. Evidence of predictable behavior of security returns. J Finance. 1990;45:881e898. https://doi.org/10.1111/j.1540-6261.1990.tb05110.x. 82. Shefrin H, Statman M. The disposition to sell winners too early and ride losers too long: theory and evidence. J Finance. 1985;40:777e790. https://doi.org/10.1111/j.1540-6261.1985.tb05002.x.

17. THANK YOU