Prediction and Analysis of Bitcoin Prices Using Diverse Regression Models

Tianze Li

2024

Abstract

Accurate prediction of cryptocurrency prices is crucial for investors and analysts due to the instability and complexity of the trading market. This study explores the effectiveness of diverse predictive models - Long Short-Term Memory (LSTM), eXtreme Gradient Boosting (XGBoost), and Random Forest (RF) - in forecasting Bitcoin prices. The inclusion of these diverse models, each representing different approaches to regression and machine learning, allows for a more comprehensive analysis of predictive accuracy. Simulations are conducted using historical Bitcoin price data from Yahoo Finance, evaluating the models based on their performance metrics: R-squared (R²) score, Mean Absolute Error (MAE), and Root Mean Squared Error (RMSE). The LSTM model demonstrated superior performance with an R² score of 0.955, an MAE of 0.04, and an RMSE of 0.0508, showing its ability to capture complex temporal dependencies. RF also performed well, achieving an R² of 0.936, MAE of 0.0459, and RMSE of 0.0604. In contrast, XGBoost lagged with an R² of 0.654, MAE of 0.1009, and RMSE of 0.1406. This study highlights the strengths of using diverse regression models for predicting Bitcoin prices, with LSTM emerging as the most effective model, while also providing insights into real-market transactions.

Paper Citation

in Harvard Style

Li T. (2024). Prediction and Analysis of Bitcoin Prices Using Diverse Regression Models. In Proceedings of the 1st International Conference on E-commerce and Artificial Intelligence - Volume 1: ECAI; ISBN 978-989-758-726-9, SciTePress, pages 286-291. DOI: 10.5220/0013215000004568

in Bibtex Style

@conference{ecai24,
author={Tianze Li},
title={Prediction and Analysis of Bitcoin Prices Using Diverse Regression Models},
booktitle={Proceedings of the 1st International Conference on E-commerce and Artificial Intelligence - Volume 1: ECAI},
year={2024},
pages={286-291},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0013215000004568},
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 and Analysis of Bitcoin Prices Using Diverse Regression Models
SN - 978-989-758-726-9
AU - Li T.
PY - 2024
SP - 286
EP - 291
DO - 10.5220/0013215000004568
PB - SciTePress