Application of Machine Learning Techniques to Predict Price Trends during Bitcoin Halving Cycles

Abstract

The rapid expansion and volatility of the cryptocurrency market has rendered Bitcoin an important investment asset. The Bitcoin halving event has had a significant impact on the market price, creating new challenges and opportunities for investors. It is often the case that traditional forecasting methods are unable to account for the non-linear and dynamic nature of the cryptocurrency market. The objective of this research is to develop a Long Short-Term Memory (LSTM) model that can predict Bitcoin price trends by analysing fluctuations around the Bitcoin halving cycle, with a particular focus on the 2024 halving event. The data collection and preprocessing phase will encompass Bitcoin trading data from 2012 to 2024. This will include daily open and close prices, high and low prices, and volume. The feature engineering process will entail the derivation of candle-stick pattern features from raw price data and the incorporation of several technical indicators, including the Simple Moving Average (SMA), Relative Strength Index (RSI), Moving Average Convergence Divergence (MACD), and Money Flow Index (MFI). Furthermore, features indicative of the number of days until the next Bitcoin halving will be included. The model will be trained to predict price trends six months after the halving in 2024.