Predicting Bitcoin Prices for the Next 30 Days Using LSTM-Based Time Series Analysis

Abstract

Particularly since Bitcoin's value and market capitalization have skyrocketed in recent years, the cryptocurrency market has attracted a lot of attention. For investors and traders, it is difficult to make wise selections due to the volatility and unpredictability of Bitcoin values. This project's goal is to forecast Bitcoin's price for the following 30 days. This prediction problem is particularly difficult due to the extreme price volatility of Bitcoin, the lack of conventional financial indications, and the complexity of the market. The price of Bitcoin for the following 30 days is predicted using a Long Short-Term Memory (LSTM) neural network in this research. The model is trained using a dataset of historical Bitcoin prices and their corresponding features such as trading volume and market capitalization. The LSTM model is trained to capture the temporal dependencies and patterns in the data, which allows it to make predictions. Previous studies have used various machine learning algorithms to predict Bitcoin prices, including ARIMA, SVM, and Random Forest. However, LSTM has shown superior performance in capturing the temporal dependencies in sequential data. The proposed LSTM model was evaluated using metrics such as Mean Squared Error (MSE) and Root Mean Squared Error (RMSE), and the results showed that the LSTM model outperforms the previous studies' models in terms of accuracy. The proposed model's predictions can assist investors and traders in making informed decisions about buying or selling Bitcoin in the future.