Performance Comparison of ARIMA, XGBoost and LSTM Models on the NASDAQ Dataset

DOI

10.2991/978-94-6239-801-6_151How to use a DOI?

Keywords

Nasdaq Composite Index; Time Series Prediction; ARIMA Model; XGBoost Model; LSTM Model; Performance Comparison

Abstract

Based on the historical trading data of the Nasdaq Composite Index, this paper constructs three prediction models—a traditional econometric model (ARIMA), a machine learning model (XGBoost), and a deep learning model (LSTM)—under a unified experimental setting on the same dataset and train/test split. The ARIMA model is trained with rolling one-step-ahead forecasting, while the XGBoost and LSTM models are trained with feature-based regression using fixed hyperparameters. Their performance on the 2023-2024 test set is evaluated with four widely used metrics: MAE, RMSE, MAPE, and R². Results show that the ARIMA model achieves the best short-term forecasting accuracy (MAPE 0.8790%, R² 0.9941), accurately reflecting short-term fluctuations. The LSTM model ranks second, capturing the overall trend well through long-term temporal dependencies. The XGBoost model is the least accurate, especially on unidirectional trends.

Copyright

© 2026 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - Ruize Ma
PY  - 2026
DA  - 2026/10/23
TI  - Performance Comparison of ARIMA, XGBoost and LSTM Models on the NASDAQ Dataset
BT  - Proceedings of the 2026 2nd International Conference on Financial Innovation and Marketing Management (FIMM 2026)
PB  - Atlantis Press
SP  - 1439
EP  - 1447
SN  - 2352-5428
UR  - https://doi.org/10.2991/978-94-6239-801-6_151
DO  - 10.2991/978-94-6239-801-6_151
ID  - Ma2026
ER  -