Hybrid Machine Learning Architectures for Short-Term Financial Forecasting Under Dynamic Market Conditions

  • Authors

    • Sri Veera Research Scholar, India. Author

    DOI:

    https://doi.org/10.67228/30716357/IJMRSE-V9I2P103

    Published 04-11-2026

  • Financial Time-Series Forecasting, Hybrid Machine Learning, Long Short-Term Memory (LSTM), Transformer, Gradient-Boosted Trees, Market-Regime Detection, Volatility Forecasting, Ensemble Learning, Walk-Forward Validation, Nonstationary Data

    Issue

    Vol. 9 No. 2 (2026)

    Section

    Articles

  • Abstract

    Financial forecasting is difficult in the short run due to the nonlinear relationships, volatility clustering, structural changes, and unpredictability in financial markets. This paper presents a hybrid machine learning approach combining Long Short-Term Memory (LSTM) networks, attention mechanism based on transformer, gradient-boosted decision trees (GBDT) and statistical volatility modelling for enhanced forecast adaptability. It takes historical prices, trading volume, technical indicators, and other market variables to process and uncover temporal patterns and shifting market regimes. The key idea is to combine individual model predictions based on market conditions in an adaptive gating mechanism, thus capturing the sequential dependency, nonlinear interaction, and risk-related information. The methodology features partitioning of chronological data, walk forward validation, benchmarking and performance metrics such as mean absolute error, root mean square error and direction accuracy. Ablation and sensitivity analysis are suggested to determine contributions of individual components. The framework facilitates research in financial forecasting, with a focus on data integrity, computational efficiency and realistic assessment. It needs to be empirically tested on a variety of assets, market regimes and out-of-sample periods.

  • References

    [1] R. F. Engle, “Autoregressive conditional heteroscedasticity with estimates of the variance of United Kingdom inflation,” Econometrica, vol. 50, no. 4, pp. 987–1007, 1982.

    [2] T. Bollerslev, “Generalized autoregressive conditional heteroskedasticity,” Journal of Econometrics, vol. 31, no. 3, pp. 307–327, 1986.

    [3] Bolla, S. R., Uddandarao, D. P., Mokashi, S. P., Mistry, H., & Mavani, C. (2025, July). AI-Powered Fraud Detection in Real-Time Payment Systems Excellent Application of ML In Fintech-Critical and Impactful Use Case. In 2025 IEEE 4th World Conference on Applied Intelligence and Computing (AIC) (pp. 1-10). IEEE.

    [4] G. E. P. Box, G. M. Jenkins, G. C. Reinsel, and G. M. Ljung, Time Series Analysis: Forecasting and Control, 5th ed. Hoboken, NJ, USA: Wiley, 2015.

    [5] S. Hochreiter and J. Schmidhuber, “Long short-term memory,” Neural Computation, vol. 9, no. 8, pp. 1735–1780, 1997.

    [6] Brahmandam, L. M. K. (2023). Migrating Mission-Critical Enterprise Workloads from On-Premises VMware to AWS: An Empirical Study of a Multi-Account Landing-Zone Reference Architecture and the Seven Rs Decision Framework. International Journal of Emerging Trends in Computer Science and Information Technology, 4(4), 231-240. https://doi.org/10.63282/3050-9246.IJETCSIT-V4I4P124

    [7] A. Vaswani et al., “Attention is all you need,” in Proc. Advances in Neural Information Processing Systems (NeurIPS), 2017, pp. 5998–6008.

    [8] G. P. Zhang, “Time series forecasting using a hybrid ARIMA and neural network model,” Neurocomputing, vol. 50, pp. 159–175, 2003.

    [9] Gantikota, S. (2024). Shift-Left Security for Decentralized Engineering Organizations: Embedding SAST, DAST, and Penetration Testing Throughout the Software Development Lifecycle in University and Research Computing Environments. International Journal of Emerging Research in Engineering and Technology, 5(4), 175-179. https://doi.org/10.63282/3050-922X.IJERET-V5I4P118

    [10] T. Fischer and C. Krauss, “Deep learning with long short-term memory networks for financial market predictions,” European Journal of Operational Research, vol. 270, no. 2, pp. 654–669, 2018.

    [11] Gadam, H., Yadava, A. K., Arora, A. S., Yachamaneni, T., & Kotadiya, U. (2025, July). Revolutionizing Wealth Management with Temporal Attention Mechanism Based Convolutional Network for Algorithmic Stock Trading Strategies. In 2025 3rd International Conference on Data Science and Network Security (ICDSNS) (pp. 1-4). IEEE.

    [12] B. Lim and S. Zohren, “Time-series forecasting with deep learning: A survey,” Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences, vol. 379, no. 2194, Art. no. 20200209, 2021.

    [13] T. Chen and C. Guestrin, “XGBoost: A scalable tree boosting system,” in Proc. 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 2016, pp. 785–794.

    [14] Sunkara, R. (2023). Cost-Optimized Energy Compliance Testing for Smart TV Streaming Devices: Achieving Milliwatt-Precision Power Measurement at Sub-One-Thousand-Dollar per Setup. American International Journal of Computer Science and Technology, 5(6), 54-59. https://doi.org/10.63282/3117-5481/AIJCST-V5I6P105

    [15] J. D. Hamilton, “A new approach to the economic analysis of nonstationary time series and the business cycle,” Econometrica, vol. 57, no. 2, pp. 357–384, 1989.

    [16] R. J. Hyndman and G. Athanasopoulos, Forecasting: Principles and Practice, 3rd ed. Melbourne, Australia: OTexts, 2021.

    [17] K. Olorunnimbe and H. Viktor, “Deep learning in the stock market—A systematic survey of practice, backtesting, and applications,” Artificial Intelligence Review, vol. 56, pp. 2057–2109, 2023.

    [18] Sandra, K. (2024). THE REGULATED BANKING AI LAKEHOUSE. INDO-CONTINENTAL ACADEMIC PUBLISHERS.

    [19] D. H. Bailey, J. Borwein, M. López de Prado, and Q. J. Zhu, “The probability of backtest overfitting,” Journal of Computational Finance, vol. 20, no. 4, pp. 39–69, 2017.

    [20] G. E. P. Box and G. M. Jenkins, “Time series analysis: Forecasting and control,” Journal of Time Series Analysis, vol. 1, no. 2, pp. 97–100, 1980.

    [21] Paruchuri, J. K. (2021). Lakehouse architecture: Unifying data lakes and data warehouses. International Journal of Computer Techniques (IJCT), 8(1), 1-15.

    [22] R. S. Tsay, Analysis of Financial Time Series, 3rd ed. Hoboken, NJ, USA: Wiley, 2010.

  • Downloads