<p>With the social economy growing at a quick pace and the stock market seeing constant developments, more and more people are voicing concerns about investing in stocks. The importance of forecasting stock values has increased in the domain of engineering's use of cognitive computing. Utilizing data-driven tactics for forecasting stock prices, investors can effectively mitigate risks and enhance profits. Investors can use projections based on historical values and textual data to make well-informed judgments about future patterns in stock prices. Stock price anticipation is a pivotal undertaking in the financial sector that has substantial consequences for traders and investors. This article presents an in-depth comparison analysis of machine learning tactics for forecasting price fluctuations in stocks. The research deploys historical stock data and diverse technical indicators. This paper presents the Gated Recurrent Unit (GRU) model for Nasdaq stock index anticipation, which is optimized by Particle swarm optimization (PSO), Biogeography-based optimization (BBO), and Moth flame optimization (MFO). Among these optimizers, MFO has the best outcomes. Compared to the GRU scheme the optimized PSO-GRU, BBO-GRU, and MFO-optimized GRU for stock forecasting has the outcomes of 0.9807, 0.9824, and 0.9904 in coefficient of determination (<InlineEquation ID="IEq1"> <InlineMediaObject> <ImageObject Color="BlackWhite" FileRef="40745_2025_616_Article_IEq1.gif" Format="GIF" Height="16" Rendition="HTML" Resolution="72" Type="Linedraw" Width="21" /> </InlineMediaObject> <EquationSource Format="TEX">${R}^{2}$</EquationSource> <EquationSource Format="MATHML"><math> <msup> <mrow> <mi>R</mi> </mrow> <mn>2</mn> </msup> </math></EquationSource> </InlineEquation>) which shows the improvement of the presented scheme as a result of its development. The criteria used to evaluate this model are mean absolute error, root mean absolute error, and <InlineEquation ID="IEq2"> <InlineMediaObject> <ImageObject Color="BlackWhite" FileRef="40745_2025_616_Article_IEq1.gif" Format="GIF" Height="16" Rendition="HTML" Resolution="72" Type="Linedraw" Width="21" /> </InlineMediaObject> <EquationSource Format="TEX">${R}^{2}$</EquationSource> <EquationSource Format="MATHML"><math> <msup> <mrow> <mi>R</mi> </mrow> <mn>2</mn> </msup> </math></EquationSource> </InlineEquation>.</p>

Stock Prices Forecasting by Using a Novel Hybrid Method Based on the MFO-Optimized GRU Network

期刊论文
发表日期：

2025年6月18日

查看全文

Xinjian Zhang,
Guanlin Liu

摘要

With the social economy growing at a quick pace and the stock market seeing constant developments, more and more people are voicing concerns about investing in stocks. The importance of forecasting stock values has increased in the domain of engineering's use of cognitive computing. Utilizing data-driven tactics for forecasting stock prices, investors can effectively mitigate risks and enhance profits. Investors can use projections based on historical values and textual data to make well-informed judgments about future patterns in stock prices. Stock price anticipation is a pivotal undertaking in the financial sector that has substantial consequences for traders and investors. This article presents an in-depth comparison analysis of machine learning tactics for forecasting price fluctuations in stocks. The research deploys historical stock data and diverse technical indicators. This paper presents the Gated Recurrent Unit (GRU) model for Nasdaq stock index anticipation, which is optimized by Particle swarm optimization (PSO), Biogeography-based optimization (BBO), and Moth flame optimization (MFO). Among these optimizers, MFO has the best outcomes. Compared to the GRU scheme the optimized PSO-GRU, BBO-GRU, and MFO-optimized GRU for stock forecasting has the outcomes of 0.9807, 0.9824, and 0.9904 in coefficient of determination ( ${R}^{2}$ $R^{2}$ ) which shows the improvement of the presented scheme as a result of its development. The criteria used to evaluate this model are mean absolute error, root mean absolute error, and ${R}^{2}$ $R^{2}$ .