TWO PREDICTION TECHNIQUES SIMULATION FOR CRYPTOCURRENCY TRADING

Authors

DOI:

https://doi.org/10.5212/Admpg.v.12.21113.008

Abstract

Marked by the high volatility in the price of cryptocurrencies, this market is responsible for the movement of trillions of dollars, with Bitcoin as its main exponent, a currency that in 2021 exceeded the value of 60 thousand dollars. Thus, several techniques already used in the stock markets, such as technical and fundamental analysis, became part of the buying and selling strategies of investors in this market. In addition, computational techniques in the field of Artificial Intelligence, such as Long Short Term Memory (LSTM) artificial neural networks, have gained prominence in forecasting the financial market. Given this scenario, this work aims to carry out a comparative analysis between the use of technical analysis and LSTM for the prediction of favorable moments for the purchase, sale and maintenance of cryptocurrencies. The result obtained by the model using technical indicators was not satisfactory, since it obtained a profit of only 3.84% during the period of 30 days. The LSTM model, on the other hand, achieved an average profit of 22.94% in the simulations.

Published

2022-11-03

How to Cite

LARROSA FURLAN, M. F. .; IGARASHI, W. .; CORRÊA IGARASHI, D. C. . TWO PREDICTION TECHNIQUES SIMULATION FOR CRYPTOCURRENCY TRADING. ADMPG Journal, [S. l.], v. 12, n. 1, 2022. DOI: 10.5212/Admpg.v.12.21113.008. Disponível em: https://revistas.uepg.br/index.php/admpg/article/view/21113. Acesso em: 22 jul. 2024.

Most read articles by the same author(s)