Academic Journal of Business & Management, 2022, 4(4); doi: 10.25236/AJBM.2022.040419.
Zenan Ji1, Yihong Zhang2
1School of Data Science, Zhejiang University of Finance & Economics, Hangzhou, 310018, China
2School of Science, North China Institute of Science & Technology, Langfang, 065201, China
Cryptocurrencies have taken an increasingly important place in the international financial markets. However, due to the complexity of the factors affecting cryptocurrencies and the high volatility of cryptocurrency prices, predicting the future trend of cryptocurrencies has been a difficult topic. In this paper, we establish a prediction model based on LightGBM, aiming to find the optimal machine learning algorithm to predict different cryptocurrencies. Firstly, we preprocessed the data to ensure data integrity and availability. Secondly, we trained the training set using the LighGBM algorithm. Then, to find the optimal LightGBM algorithm for different cryptocurrencies, we optimized the algorithm using Grid Search. Finally, we evaluated the results by Root Mean Squared Error (RMSE). It shows that the Root Mean Squared Error of the prediction results for all cryptocurrencies is below 8%, which means LightGBM can be a good choice for monitoring cryptocurrencies.
Cryptocurrency; Residual log-returns; LightGBM; Grid Search
Zenan Ji, Yihong Zhang. Predictive Analysis of Cryptocurrency Residual Log-returns Based on Lightgbm. Academic Journal of Business & Management (2022) Vol. 4, Issue 4: 96-100. https://doi.org/10.25236/AJBM.2022.040419.
[1] Guo Yan,Wang Lirong,Han Yan. Blockchain technology in financial market:scenario application and value outlook [J]. Technology Economics, 2017, 36(07): 110-116.
[2] Li Kangzhen, Wang Hao. Research on the use of digital cryptocurrency crimes and its detection and prevention revelation [J]. Journal of Hunan Police Academy, 2018, 30(03): 57-66.
[3] Zeng Fancheng. Research on cryptocurrency price prediction based on deep learning method and quantile regression [D]. Jiangxi University of Finance and Economics, 2021.
[4] S. Saadah and A. A. Ahmad Whafa, "Monitoring Financial Stability Based on Prediction of Cryptocurrencies Price Using Intelligent Algorithm," 2020 International Conference on Data Science and Its Applications (ICoDSA), 2020, pp. 1-10, doi: 10.1109/ICoDSA50139.2020.9212968.
[5] Gullapalli S. Learning to predict cryptocurrency price using artificial neural network models of time series [J]. 2018.
[6] Sovbetov Y. Factors influencing cryptocurrency prices: Evidence from bitcoin, ethereum, dash, litcoin, and monero [J]. Journal of Economics and Financial Analysis, 2018, 2(2): 1-27.
[7] Valencia F, Gómez-Espinosa A, Valdés-Aguirre B. Price movement prediction of cryptocurrencies using sentiment analysis and machine learning [J]. Entropy, 2019, 21(6): 589.
[8] A. Inamdar, A. Bhagtani, S. Bhatt and P. M. Shetty, "Predicting Cryptocurrency Value using Sentiment Analysis," 2019 International Conference on Intelligent Computing and Control Systems (ICCS), 2019, pp. 932-934, doi: 10.1109/ICCS45141.2019.9065838.