Stock Price Prediction Based on Linear Regression and Significance Analysis

Shengzhou Li

2025

Abstract

Accurate stock price prediction is fundamental to financial market efficiency, enabling informed trading strategies and systemic risk mitigation in increasingly volatile global markets. To address the critical yet underexplored issue of feature selection efficacy, this paper investigates stock price prediction for Apple Inc. (AAPL) over the 2013–2018 period by applying a linear regression model and analyzing four fundamental price features (open, high, low, close) along with their five-day lags. Single-group and combined-group approaches were applied to forecast next-day and five-day-ahead closing prices. The aim was to clarify which feature combination offers greater predictive benefit. Results show that while a single-group model relying on closing prices alone performs relatively well, its accuracy does not significantly differ from the combined model. This finding suggests that feature redundancy may reduce potential gains in short-term contexts. Meanwhile, the partial F-test indicates that high price features exhibit notable statistical significance for capturing market peaks and volatility, whereas information from open, low, and close can be partially overlapped by other variables.

Paper Citation

in Harvard Style

Li S. (2025). Stock Price Prediction Based on Linear Regression and Significance Analysis. In Proceedings of the 2nd International Conference on Data Science and Engineering - Volume 1: ICDSE; ISBN 978-989-758-765-8, SciTePress, pages 596-603. DOI: 10.5220/0013702600004670

in Bibtex Style

@conference{icdse25,
author={Shengzhou Li},
title={Stock Price Prediction Based on Linear Regression and Significance Analysis},
booktitle={Proceedings of the 2nd International Conference on Data Science and Engineering - Volume 1: ICDSE},
year={2025},
pages={596-603},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0013702600004670},
isbn={978-989-758-765-8},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 2nd International Conference on Data Science and Engineering - Volume 1: ICDSE
TI - Stock Price Prediction Based on Linear Regression and Significance Analysis
SN - 978-989-758-765-8
AU - Li S.
PY - 2025
SP - 596
EP - 603
DO - 10.5220/0013702600004670
PB - SciTePress