Comparative Analysis of Regression Models for Stock Price Prediction: LSTM, ARIMA, SVM

Himani Deshpande, Vikas Talreja, Muskan Tolani, Harshvardhan Rijhwani, Rohit Sharma

2024

Abstract

The research deals with a critical and challenging issue in the dynamic financial field. It critically evaluates and predicts stock prices using three popular regression models: Long Short-Term Memory (LSTM), Autoregressive Integrated Moving Average (ARIMA), and Support Vector Machine (SVM). Using a rich dataset that spans market volatility over long-term trends, short-term fluctuations, and an unprecedented period when COVID-19 struck, the research tries to determine which model gives the most accurate forecast for stock prices. The data was meant to cover the economic giants such as HDFC, ONGC, Tata, and Adani to give relevance and comprehensiveness to the study. This study gives an insight into time-based stock price analysis. The findings are very helpful to the financial expert in that they provide critical insights helpful in choosing the appropriate model based on the needs of the person carrying out analysis and thus aid in forecast accuracy. Experimental analysis suggests that, among the selected methods, the ARIMA model has given the highest prediction accuracy, which is approximately 95.26%. MSE and RMSE for the model come out to be 1.355 and 1.164 for Adani Ports, respectively, hence proving the model's performance to be very good even on long-run datasets. Further, ARIMA performance on a short-run dataset, for HDFC, and on ONGC for a novel COVID-19 set cements further that strength. Such practical evidence places ARIMA on the most reliable procedure while walking through the ambiguity of financial market forecasting, providing financial analysts with a very effective tool for strategic decision-making. Thus concludes that ARIMA helps to add value to the predictive models and promotes strategic decisions in stock markets through forecasting.

Paper Citation

in Harvard Style

Deshpande H., Talreja V., Tolani M., Rijhwani H. and Sharma R. (2024). Comparative Analysis of Regression Models for Stock Price Prediction: LSTM, ARIMA, SVM. In Proceedings of the 1st International Conference on Cognitive & Cloud Computing - Volume 1: IC3Com; ISBN 978-989-758-739-9, SciTePress, pages 215-221. DOI: 10.5220/0013307600004646

in Bibtex Style

@conference{ic3com24,
author={Himani Deshpande and Vikas Talreja and Muskan Tolani and Harshvardhan Rijhwani and Rohit Sharma},
title={Comparative Analysis of Regression Models for Stock Price Prediction: LSTM, ARIMA, SVM},
booktitle={Proceedings of the 1st International Conference on Cognitive & Cloud Computing - Volume 1: IC3Com},
year={2024},
pages={215-221},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0013307600004646},
isbn={978-989-758-739-9},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 1st International Conference on Cognitive & Cloud Computing - Volume 1: IC3Com
TI - Comparative Analysis of Regression Models for Stock Price Prediction: LSTM, ARIMA, SVM
SN - 978-989-758-739-9
AU - Deshpande H.
AU - Talreja V.
AU - Tolani M.
AU - Rijhwani H.
AU - Sharma R.
PY - 2024
SP - 215
EP - 221
DO - 10.5220/0013307600004646
PB - SciTePress