Year: 2024
Type: article
Abstract: This research explores the application of Long Short-Term Memory (LSTM) networks for short-term Bitcoin price prediction, addressing the need for reliable models due to Bitcoin's high volatility and trading volume. The study employs historical data from Kaggle to predict the direction and magnitude of price changes within a five-minute interval. Implementation includes preprocessing the data, normalizing prices, and generating sequences for LSTM input. Two LSTM models were developed: one for directional prediction and another for magnitude. Training results showed a directional accuracy of approximately 75.10%, demonstrating the feasibility of LSTM networks for financial forecasting and contributing to Bitcoin price prediction research, setting the stage for future real-time applications.
Author: Kannan Balasubramanian
Institution: SASTRA University
Language: English
Field: Decision Sciences
Domain: Social Sciences
Open Access status: hybrid
OpenAlex ID: W4404823633