A Hybrid Neuro-Fuzzy and Deep Learning Framework for Stock Price Forecasting in Emerging Markets

Precise forecasting of stock prices remains a tenacious challenge due to the inherent uncertainty, nonlinearity, and instability of financial markets, particularly in emerging economies. This paper proposes a Hybrid Neuro-Fuzzy and Deep Learning (HNFDL) Framework that combines fuzzy inference systems with deep learning architectures to improve stock price forecasting accuracy and interpretability. The model leverages four technical indicators, Relative Strength Index, Moving Average Convergence Divergence, Stochastic Oscillator and On-Balance Volume, as primary inputs into a fuzzy decision layer. To address fuzzy rule optimization and parameter tuning, genetic algorithms (GA) and long short-term memory (LSTM) networks are incorporated, enabling the system to learn temporal dependencies while upholding linguistic interpretability. Empirical validation was executed using historical data from Dangote Cement Plc, Zenith Bank Plc and the Nigerian Stock Exchange All-Share Index. Unlike previous studies relying on static splits, this study employs Walk-Forward Validation to minimize look-ahead bias. Crucially, an ablation study validated the necessity of the Genetic Algorithm, demonstrating an 8.4% reduction in Root Mean Square Error (RMSE) compared to a non-optimized hybrid baseline. The full HNFDL model achieved a directional accuracy of 68.4%, outperforming conventional LSTM and Fuzzy-only methods. Importantly, the superiority of the model was confirmed using the Diebold-Mariano test (p < 0.01), proving that enhancements are statistically significant and not due to random noise. The study demonstrates the importance of combining explainable fuzzy reasoning with adaptive deep neural architectures to enhance decision-making confidence among investors. Supplementary Material File (journal.docx) - Download - 2.73 MB Information & Authors Information Version history Copyright This work is licensed under a Non Exclusive No Reuse License.

Authors Metrics & Citations Metrics Article Usage 193views 75downloads Citations Download citation Olumide Sunday Adewale, Emmanuel Onwuka Ibam, Johnson Bisi Oluwagbemi. A Hybrid Neuro-Fuzzy and Deep Learning Framework for Stock Price Forecasting in Emerging Markets. Authorea. 26 November 2025. DOI: https://doi.org/10.22541/au.176418594.47960497/v1 DOI: https://doi.org/10.22541/au.176418594.47960497/v1 If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download. For more information or tips please see 'Downloading to a citation manager' in the Help menu.