Forecasting returns volatility of cryptocurrency by applying various deep learning algorithms

Abstract The study aims at forecasting the return volatility of cryptocurrencies using several machine learning algorithms, like neural network autoregressive (NNETAR), cubic smoothing spline (CSS), and group method data handling (GMDH-NN) algorithm. used in this is spanning from April 14, 2017, to October 30, 2020, covering 1296 observations. We predict four cryptocurrencies, namely Bitcoin, Ethereum, XRP, Tether, compare their predictive power terms accuracy. capabilities CSS, NNETAR, GMDH-NN are compared evaluated by mean absolute error (MAE) root-mean-square (RMSE). Regarding Bitcoin XRP markets, forecasted results remarkably suggest that contrast rival approaches, CSS can be an effective model boost predicting accuracy sense it has lowest forecast errors. Considering Ethereum markets’ volatility, MAE RMSE associated with NNETAR smaller than algorithm, which ensures effectiveness as competing approaches. Similarly, case Tether corresponding reveal algorithm efficient technique enhance performance. notice no single tool performed uniformly for all cryptocurrency markets. policymakers adopt accordingly.