Artificial Intelligence Application of Back-propagation Neural Network in Cryptocurrency Price Prediction

Muhammad Sahi (Universitas Islam Negeri Maulana Malik Ibrahim Malang)
Galan Ramadan Harya Galib (Universitas Islam Negeri Maulana Malik Ibrahim Malang)

International Journal of Electrical and Intelligent Engineering
Vol 1, No 1 (2025)

This study explores the use of Deep Learning and Artificial Intelligence (AI), particularly Artificial Neural Networks (ANN), for cryptocurrency price prediction. Given the high volatility of crypto markets, traditional models often underperform. A backpropagation-based ANN with a 7-5-1 architecture is proposed and tested using historical Bitcoin data. The model achieves high accuracy, with a Mean Squared Error (MSE) of 4.0431e-04, equivalent to 99.96% accuracy, demonstrating its ability to capture complex nonlinear patterns. However, overfitting remains a concern, emphasizing the need for robust generalization and feature selection. The results validate the potential of ANN in crypto forecasting and encourage further research using diverse features and assets.