Document Type : Original Article

Authors

• Kamran Pakizeh ¹
• Arman Malek ¹
• Mahya Karimzadeh khosroshahi ¹
• Hasan Hamidi Razi ²

¹ Faculty of Financial Sciences, Kharazmi University, Tehran, Iran

² Department of Hydraulics Engineering, Tarbiat Modares University, Tehran, Iran

Abstract

Cryptocurrencies, which are digitally encrypted and decentralized,

continue to attract attention of financial market players across

the world. Because of high volatility in cryptocurrency market, predicting

price of cryptocurrencies has become one of the most complicated

fields in financial markets. In this paper, we use Long Short-Term Memory

(LSTM) and Gated Recurrent Unit (GRU) models to predict price

of four well-known cryptocurrencies of Bitcoin (BTC), Ethereum (ETH),

Litecoin (LTC), and Ripple (XRP). These models are subdivisions of Artificial

Intelligence, machine learning and data science. The main aim of

this paper is to compare the accuracy of above-mentioned models in forecasting

time series data, to find out which model can better predict price

in these four cryptocurrencies. 43 variables consisting of 28 technical

indicators and t+10 lags were calculated and appended to the Open,

High, Low, Close and Volume (OHLCV) data for selected cryptocurrencies.

Applying random forest as feature selection, 25 variables were

chosen, 24 of them selected as feature (independent variables) and one

as a dependent variable. Each attribute value was converted into a relative

standard score, followed by Min-max scaling; we compare models

and results of Dieblod Mariano test that is used to examine whether the

differences in predictive accuracy with these two models are significant,

reveal that LSTM reaches better accuracy than GRU for BTC and ETH,

but both models convey the same accuracy for LTC and XRP.

Keywords

• Cryptocurrency
• Long short-term memory
• Gated recurrent unit
• Random forest classifier