World Summit on Management Sciences
  • Follow

Accepted Abstracts

A Gated Recurrent Unit Approach to Bitcoin Price Prediction

Saket Kumar* 
Reserve Bank of India, India

Citation: Kumar S (2020) A Gated Recurrent Unit Approach to Bitcoin Price Prediction. SciTech Management Sciences 2020. Thailand 

Received: February 17, 2020         Accepted: February 19, 2020         Published: February 19, 2020


In today’s era of big data, deep learning and artificial intelligence are changing the facades of traditional finance. The machine learning methods have now become the new way forward for financial modelling, optimization and even trading. The financial world has also from time to time discovered new assets to invest and one such new asset which is storming the eyes of the market is the crypto currency market. Since the crypto market is still evolving a well-documented study and research on the probable factors driving the markets based on traditional finance theories of efficient markets, factor models etc. is still under-way. The price and volatility prediction based on time-series models have been extensively done. Out of nearly 6000 crypto’s trading today, Bitcoin’s rise has been most astounding and is the most used crypto today. Researchers have investigated various state of the art machine learning models to predict Bitcoin price and volatility. Machine learning models like recurrent neural network (RNN) and long short-term memory (LSTM) have been shown to perform better than traditional time series models in crypto currency price prediction. However, very few studies have applied sequence models with robust feature engineering to predict future pricing. In this study, we investigate a framework with a set of advanced machine learning forecasting methods with a fixed set of exogenous and endogenous factors to predict daily Bitcoin prices. We study and compare different approaches using the root mean squared error (RMSE). Experimental results show that the gated recurring unit (GRU) model with recurrent dropout performs better than popular existing models. We also show that simple trading strategies, when implemented with our proposed GRU model and with proper learning, can lead to financial gain.