Algorithmic Trading Model Development for BTCUSDT Crypto Market

Algorithmic Trading Model Development for BTCUSDT Crypto Market

ABSTRACT

This project explores the application of advanced time-series forecasting techniques to predict future Bitcoin prices. Using a dataset of daily Bitcoin prices from 2018 to 2024, the study evaluates several machine learning and statistical models—LSTM, ARIMA, GRU, and Prophet—to forecast price movements. Each algorithm brings unique strengths: LSTM and GRU are designed to capture long-term dependencies in sequential data, ARIMA is a classic statistical approach for handling seasonality and trends, and Prophet offers flexibility and interpretability in forecasting. By comparing their predictive performance, the project provides insights into the most effective methodologies for anticipating cryptocurrency price fluctuations. This research ultimately contributes to a deeper understanding of how data-driven approaches can aid decision-making in volatile markets.