Data-Driven Algorithmic Trading with Market Sentiment Insights

• Onyeka Alimele

  Author

Cryptocurrency markets are experiencing exponential growth, creating the need for a good predictive model. In particular, this model must be able to capture the volatility and highly non-linear behaviour of crypto prices. The present study suggests a framework that is fusion-based and is hybrid in nature, using both the indicators of technical analysis and the various classical supervised ML algorithms, which help in the prediction of the price of cryptocurrency. The method detailed in this paper uses historical OHLC data for five cryptocurrencies: Ripple (XRP), Ethereum (ETH), Litecoin (LTC), Cardano (ADA), and Polkadot (DOT). The study makes use of accepted technical indicators that comprise the relative strength index (RSI), moving average convergence divergence (MACD), and Bollinger bands to perform feature engineering with these indicators, later used as the input for SVM, decision tree (DT), and random forest (RF) models. The models were trained featuring a validation procedure accounting for time series effects. The mean squared error (MSE), root mean squared error (RMSE), mean absolute error (MAE), mean absolute percentage error (MAPE), and coefficient of determination ($R^2$). The existing experimental evidence suggests that Random Forest has less prediction error and better generalization ability than SVM and Decision Tree. A combination of indicators using ensemble learning can improve the forecasting accuracy of the price of cryptocurrency.

• PDF

Copyright (c) 2026 International Journal of Adaptive Management and Business Intelligence

This work is licensed under a Creative Commons Attribution 4.0 International License.