Application of Inverse Optimization Algorithms in Neural Network Models for Short-Term Stock Price Forecasting
Open Access
- 9 September 2025
- journal article
- research article
- Published by MDPI AG in Big Data and Cognitive Computing
- Vol. 9 (9) , 235
Abstract
This paper introduces novel inverse optimization algorithms (RC and DC) for neural network training in stock price forecasting in an attempt to overcome the traditional gradient descent limitation of local minima convergence. The key novelty is a stochastic algorithm for inverse problems adapted to neural network training, where target function values decrease iteratively through selective weight modification. Experimental analysis used closing price data from 40 Russian companies, comparing traditional activation functions (linear, sigmoid, tanh) with specialized functions (sincos, cloglogm, mish) across perceptrons and single-hidden-layer networks. Key findings show the superiority of the DC method for single-layer networks, while RC proves most effective for hidden-layer networks. The linear activation function with the RC algorithm delivered optimal results in most experiments, challenging conventional nonlinear activation preferences. The optimal architecture, namely, a single hidden layer with two neurons, achieved the best prediction accuracy in 70% of cases. The research confirms that inverse optimization algorithms can provide higher training efficiency than classical gradient methods, offering practical improvements for financial forecasting.Keywords
Funding Information
- TUSUR Development Program for 2025–2036 of the Strategic Academic Leadership Program “Priority 2030”
This publication has 40 references indexed in Scilit:
- A dissolved oxygen levels prediction method based on single-hidden layer feedforward neural network using neighborhood information metricApplied Soft Computing, 2024
- Stock picking with machine learningJournal of Forecasting, 2023
- A modified Adam algorithm for deep neural network optimizationNeural Computing and Applications, 2023
- Application of Meta-Heuristic Algorithms for Training Neural Networks and Deep Learning Architectures: A Comprehensive ReviewNeural Processing Letters, 2022
- Genetic Algorithms in the Fields of Artificial Intelligence and Data SciencesAnnals of Data Science, 2021
- Gradient based invasive weed optimization algorithm for the training of deep neural networkMultimedia Tools and Applications, 2021
- A survey of swarm and evolutionary computing approaches for deep learningArtificial Intelligence Review, 2019
- Comparison of new activation functions in neural network for forecasting financial time seriesNeural Computing and Applications, 2010
- Threshold heteroskedastic modelsJournal of Economic Dynamics and Control, 1994
- Generalized autoregressive conditional heteroskedasticityJournal of Econometrics, 1986