Prediction of DOGE Based on Random Forest, Long Short-Term Memory and Transformer

Shengjie Yu

2024

Abstract

As cryptocurrencies have surged in value and importance in recent years, Dogecoin has been increasingly regarded as an investment asset. Due to its high volatility, the demand to forecast Dogecoin prices using machine learning techniques is rising. This study explores the application of four models, i.e., Linear Regression, Random Forest, Long Short-term Memory (LSTM), and Transformer, in forecasting the hourly prices of Dogecoin. Through comprehensive experiments, using MAE, MSE, and R-squared data as test standards, the LSTM model demonstrated superior performance, achieving the lowest error rates compared to the other models, followed by the linear regression model. The Random Forest model also performed reasonably well but fell short of the linear regression model. The Transformer model, despite its advanced architecture, delivered the poorest performance, highlighting its limitations in this specific time series forecasting task. These findings suggest that LSTM models may be more effective for time series prediction tasks in the cryptocurrency market, highlighting the need for further research into advanced machine learning techniques for financial forecasting.

Download


Paper Citation


in Harvard Style

Yu S. (2024). Prediction of DOGE Based on Random Forest, Long Short-Term Memory and Transformer. In Proceedings of the 1st International Conference on E-commerce and Artificial Intelligence - Volume 1: ECAI; ISBN 978-989-758-726-9, SciTePress, pages 188-194. DOI: 10.5220/0013212500004568


in Bibtex Style

@conference{ecai24,
author={Shengjie Yu},
title={Prediction of DOGE Based on Random Forest, Long Short-Term Memory and Transformer},
booktitle={Proceedings of the 1st International Conference on E-commerce and Artificial Intelligence - Volume 1: ECAI},
year={2024},
pages={188-194},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0013212500004568},
isbn={978-989-758-726-9},
}


in EndNote Style

TY - CONF

JO - Proceedings of the 1st International Conference on E-commerce and Artificial Intelligence - Volume 1: ECAI
TI - Prediction of DOGE Based on Random Forest, Long Short-Term Memory and Transformer
SN - 978-989-758-726-9
AU - Yu S.
PY - 2024
SP - 188
EP - 194
DO - 10.5220/0013212500004568
PB - SciTePress