• Version
• Download 16
• File Size 1.39 MB
• File Count 1
• Create Date 19 March 2026
• Last Updated 19 March 2026

Predictive Analysis of Bitcoin Price Trends Usingsupervised Machine Learning Algorithms

Dr. Y. Mohammed Iqbal1, T. Dhanushkumar2, Dr. S. Peerbasha3,Dr. M. Mohamed Surputheen4, Dr. M.
Rajakumar5
Department of Computer Science, Jamal Mohamed College, Affiliated to Bharathidasan University,
Tiruchirappalli, Tamil Nadu, India

Abstract- Cryptocurrency markets, particularly Bitcoin,are characterized by high volatility and complex non linear price movements, making trend prediction a significant challenge. The purpose of this research is to compare six different supervised machine learning models: Logistic Regression, Decision Tree, Random Forest, K-Nearest Neighbors (KNN), XGBoost, and Support Vector Machine (SVM), using a large dataset ofdaily changes in Bitcoin prices from September 2014  through January 2026 (24,863 records). The dataset is augmented by twelve different technical indicators (e.g. RSI, MACD, and multiple Simple Moving Averages (SMA)) that were created as input variables. As thefinancial data in this dataset is temporal in nature, time  series cross-validation (Time Series Split) was used to evaluate the models in order to reduce the likelihood ofoverfitting due to random sample shuffling. Based on the  experimental results, the Random Forest and XGBoost ensemble models are significantly better at predicting the price change for cryptocurrencies than the non-ensemble models, with the Random Forest model exhibiting an accuracy rate of 79.18% and AUC rate of 0.8675 for the validation folds, while the XGBoost model exhibited a 74.66% accuracy rate. This implies that advanced tree based ensemble models are capable of providing asignificant degree of prediction for cryptocurrency price trends if they are properly regularized and validated against financial market noise. Additionally, the predictive superiority of the ensemble models over the non-ensemble models was demonstrated statisticallythrough the use of McNemar's test and Point-Biserialcorrelation (p < 0.05).
Keywords: Bitcoin, Machine Learning, Technical Analysis, Random Forest, XGBoost, Price Prediction, Financial Forecasting, Cryptocurrency, Supervised Machine Learning, Ensemble Learning, Time Series Cross-Validation, Feature Engineering.