BITCOIN PRICE PREDICTION USING LSTM

Bitcoin, the most prominent cryptocurrency, exhibits high price volatility, and predicting the price is a serious challenge for investors and traders.Recurrent neural networks, specifically Long Short-Term Memory (LSTM) networks, have become a robust means of time-series forecasting, especially for forecasting Bitcoin prices.This systematic review examines developments in LSTM-based models for predicting Bitcoin prices, including their architecture, data preprocessing methods, and performance evaluation metrics.The article combines insights from current research, emphasizing the advantages of LSTM in extracting temporal relationships in highly volatile cryptocurrency markets.Important applications, like trading strategy optimization and risk management, are explained along with limitations, such as data noise, computational intensity, and model overfitting.The review also touches on future research directions, such as incorporating sentiment analysis and hybrid models to improve prediction accuracy and resilience.