From Past to Present: A Historical Review of Recurrent Neural Networks in Bitcoin Price Prediction

This comprehensive review synthesizes insights from many research papers on Bitcoin, encompassing aspects like its foundational peer-to-peer system and the application of advanced machine learning for price prediction. Examining methodologies such as Bayesian neural networks and LSTM models, the paper identifies trends, challenges, and advancements in understanding Bitcoin prices. Additionally, it intersects with Indonesia’s monetary system, employing diverse predictive mod- eling techniques. The study evaluates recurrent neural networks and LSTM algorithms for forecasting, delves into Bitcoin’s architectural assessment using Systems Modeling Language, and compares machine learning models. Spanning two decades of econophysics, it explores milestones, challenges, and addresses Bitcoin and Ethereum’s role as safe-havens during the global COVID 19 pandemic.