Comparing neural network with linear regression for stock market prediction

Kurniawan, Fachrul ORCID: https://orcid.org/0000-0002-3709-8764, Arif, Yunifa Miftachul, Nugroho, Fresy and Ikhlayel, Mohammed (2023) Comparing neural network with linear regression for stock market prediction. Bulletin of Social Informatics Theory and Application, 7 (1). pp. 8-13. ISSN 2614-0047

Preview

Text 15132.pdf - Published Version Available under License Creative Commons Attribution Share Alike. Download (622kB) | Preview

Abstract

There are both gains and losses possible in stock market investing. Brokerage firms' stock investments carry a higher risk of loss since their stock prices are not being tracked or analyzed, which might be problematic for businesses seeking investors or individuals. Thanks to progress in information and communication technologies, investors may now easily collect and analyze stock market data to determine whether to buy or sell. Implementing machine learning algorithms in data mining to obtain information close to the truth from the desired objective will make it easier for an individual or group of investors to make stock trades. In this study, we test hypotheses on the performance of a financial services firm's stock using various machine learning and regression techniques. The relative error for the neural network method is only 0.72 percentage points, while it is 0.78 percentage points for the Linear Regression. More training cycles must be applied to the Algortima neural network to achieve more accurate results.

Item Type:	Journal Article
Keywords:	neural network; linear regression; stock market; prediction
Subjects:	08 INFORMATION AND COMPUTING SCIENCES > 0806 Information Systems > 080605 Decision Support and Group Support Systems 08 INFORMATION AND COMPUTING SCIENCES > 0803 Computer Software 08 INFORMATION AND COMPUTING SCIENCES > 0806 Information Systems
Divisions:	Faculty of Technology > Department of Informatics Engineering
Depositing User:	Fachrul Kurniawan
Date Deposited:	23 Jun 2023 08:40

Downloads

Origin of downloads

Actions (login required)

View Item