Bitcoin Price Forecasting: A Comparison of LSTM and Feedforward Neural Network

Abstract

Technology and finance have experienced a change which leads to the rise of cryptocurrencies, with Bitcoin serving as a pioneer. Investors, researchers, and fans all share a fascination with bitcoin because of its decentralized structure and cryptographic security. The price volatility of Bitcoin has attracted attention and presents opportunities as well as difficulties for traders and analysts. For navigating this turbulent market, precise price prediction models are essential. In order to anticipate Bitcoin prices, this work compares Long Short-Term Memory (LSTM) with Feedforward Neural Networks (FFN). The work assesses the prediction ability of these two neural network architectures using historical pricing data and maybe other relevant factors. The comparison covers issues with anticipating Bitcoin price movements' accuracy, resilience, and generalizability.