Forecast of the optimal activation function for the sta

Abstract

Stablecoins are a special case of cryptocurrencies which emerged as a response to the high volatility generated in the cryptocurrency market, using a pegging mechanism the stablecoins obtain more stability that could be simulated using artificial neural network (ANN). The objective of this paper is to determine the best activation function between sigmoid, linear or tanh and to forecast a period of 4 days using a basic ANN that could help an investor to determine an entrance or exit point in a stablecoin. As a main result, in all cases the sigmoid activation function is the best option using ANN, although using a basic ANN doesn’t present an accurate forecast in all cases; therefore, it could be used as a supplementary tool that would help the investor to forecast the stablecoin market and improve their portfolio. This was obtained by comparing statistically 50 registers obtained of root-mean-square-error (RMSE).

Stablecoins  | Activation function | Artificial neural network | Forecast 

References

1. Atsalakis G. S. , Ioanna G. Atsalaki , Fotios Pasiouras, Constantin Zopounidis , Bitcoin price forecasting with neuro-fuzzy techniques, European Journal of Operational Research (2019), doi: https://doi.org/10.1016/j.ejor.2019.01.040
   
2. BBC News Mundo (2021) Bitcoin: El Salvador, el primer país del mundo en autorizar que la criptomoneda sea de curso legal. Available online: https://www.bbc.com/mundo/noticias-america-latina-57373067 (Accessed on June 29 2021)
   
3. Blockchain (2019) The State of Stablecoins. Technical Report. https://www.blockchain.com/ru/static/pdf/StablecoinsReportFinal.pdf.
   
4. Bottou L. (2010) Large-Scale Machine Learning with Stochastic Gradient Descent. Proceedings in Computational Statistics, 19, 177-186. DOI 10.1007/978-3-7908-2604-3 16
   
5. Buffet W. (2020) Available online https://www.cnbc.com/2020/02/24/warren-buffett-cryptocurrency-has-no-value.html (accessed on April 4 2021)
   
6. CNBC (2017) Should you invest in cryptocurrency? | CNBC Explains [TV]. From https://www.youtube.com/watch?v=3cCUkmTJNo8
   
7. CNBC (2021) Why central banks want to launch digital currencias | CNBC Explains [TV]. From https://www.cnbc.com/video/2021/02/11/why-central-banks-want-to-launch-digital-currencies.html
   
8. Coinmarketcap.com (2021) Cryptocurrencies Prices, Charts And Market Capitalizations | CoinMarketCap Available online: https://coinmarketcap.com/ (Accessed on April 3 2021)
   
9. Hecht A. (March 11 2021). USD Coin, Tether: 2 Stablecoins Tracking the U.S. Dollar; Which Will Survive? Investing.com Recovered from https://www.investing.com/analysis/usd-coin-tether-2-stablecoins-tracking-the-us-dollar-which-will-survive-200566521
   
10. Ketkar N. (2017) Deep Learning with Python: A Hands-on Introduction. Bangalore, Karnataka: Apress Media LLC
    
11. Klages-Mundt et al. (2020) Stablecoins 2.0: Economic Foundations and Risk-based Models. 2nd ACM Conference on Advanced in Financial Technologies (AFT’20), 59 – 79, https://doi.org/10.1145/3419614.3423261
    
12. Kristjanpoller W. and Minutolo M. C. (2017) A hybrid volatility forecasting framework integrating GARCH, Artificial Neural network, Technical Analysis and Principal Components Analysis, Expert Systems With Applications (2018), doi: 10.1016/j.eswa.2018.05.011 Lahmiri S. and Bekiros S. (2018) Cryptocurrency forecasting with deep learning chaotic neural networks. Chaos, Solitons and Fractals 118, 35-40. https://doi.org/10.1016/j.chaos.2018.11.014
    
13. Lahmiri S. and Bekiros S. (2019) Intelligent forecasting with machine learning trading systems in chaotic intraday Bitcoin market. Chaos, Solitons and Fractals 133, 109641. https://doi.org/10.1016/j.chaos.2020.109641
    
14. Lahmiri S., Saade R. G., Morin D. and Nebede F. (2020) An Artificial Neural Network Based Ensemble System to Forecast Bitcoin Daily Trading Volume. 5th International Conference on Cloud Computing and Artificial Intelligence: Technologies and Applications (CloudTech). DOI: 10.1109/CloudTech49835.2020.9365913
    
15. Lahmiri S. and Bekiros S. (2021) Deep Learning Forecasting in Cryptocurrency High-Frequency Trading. Springer Science and Business Media, LLC. https://doi.org/10.1007/s12559-021-09841-w
    
16. McCaleb J., Lin Lindsay and Lund J. (2018). Programmable money Will central banks take the lead? IBM Institute for Business Value. Available online https://www.ibm.com/downloads/cas/WVJNWYO4 (accessed May 1 2021)
    
17. McNally S. et al. (2018) Predicting the Price of Bitcoin Using Machine Learning. 26th Euromicro International Conference on Parallel. 10.1109/PDP2018.2018.00060
    
18. Nakano M., Takahashi A. and Takahashi S. (2018) Bitcoin technical trading with artificial neural network. Physica A 510, 587-609. https://doi.org/10.1016/j.physa.2018.07.017
    
19. Nakano M. and Takahashi A. (2020) A new investment method with AutoEncoder: Application to cryptocurrencies. Experts systems with Applications 162, 113730. https://doi.org/10.1016/j.eswa.2020.113730
    
20. Sashiko S. R. O. and Flores de la Mota I. (2019). Complex network analysis: Mexico’s City metro system. European Modeling and Simulation Symposium, 2019
    
21. Tasarova et al. (2020). Accounting and features of mathematical modeling of the system to forecast cryptocurrency exchange rate. Growing Science Accounting (6) 357-364, 10.5267/j.ac.2020.1.003
    
22. Wang G. J. et al. (2020). Are stablecoins truly diversifiers, hedges, or safe havens against traditional cryptocurrencies as their name suggests? Elsevier B.V. (54) https://doi.org/10.1016/j.ribaf.2020.101225