Facebook pixel tracking

The Science and Information (SAI) Organization publishes open-access peer-reviewed journals in computer science and artificial intelligence.

Contact Info
Website thesai.org
Follow Us
Contact Info
Follow Us
Research Article | Open Access |

Feature Selection for Machine Learning and Deep Learning-Based Heart Disease Prediction: A Systematic Review of Predictive Performance and Evaluation Practices

Author 1: Wan Aezwani Wan Abu Bakar Author 2: Mustafa Man Author 3: Nurnadiah Zamri Author 4: Ily Amalina Ahmad Sabri Author 5: Noor Suhana Sulaiman Author 6: Azliza Yacob Author 7: Nor Sukinah Aziz Author 8: Sharif Ahmad Author 9: Nur Laila Najwa Josdi
International Journal of Advanced Computer Science and Applications (IJACSA) · Vol. 17, No. 10 · Published 2026

DOI: 10.14569/IJACSA.2026.0171050

Abstract

The persistent global burden of cardiovascular disease has driven growing interest in machine-learning (ML) and deep-learning (DL) approaches for heart disease prediction. Structured clinical datasets often contain redundant or weakly informative variables, prompting the integration of feature-selection techniques into predictive pipelines. However, whether feature selection improves prediction on its own, rather than merely accompanying other pipeline changes, remains unclear. This systematic review, conducted following PRISMA 2020 guidelines, synthesizes evidence from 72 studies published between 2021 and 2026 across conventional ML classifiers and neural architectures such as Multilayer Perceptrons (MLPs). Using a PECO framework and a purpose-built five-level attribution hierarchy, the review assessed studies not only for reported predictive outcomes but also for how consistently they held classifier configuration, training procedures, and validation designs constant when comparing full and reduced feature sets. Furthermore, this review establishes a precise clinical taxonomy distinguishing cardiovascular disease, coronary heart disease, heart failure, and myocardial infarction. Dimensionality reduction emerged as a consistent outcome of feature selection, but its effect on predictive performance varied considerably by dataset, classifier, and selection strategy. Few studies met the higher tiers of the attribution hierarchy, where classifier and training configurations are fixed, and feature selection is confined to training data to avoid leakage. This shortfall was particularly evident for Recursive Feature Elimination (RFE), which has an established evidence base with tree-based estimators but limited evidence for standalone MLP architectures under controlled, leakage-safe conditions. External validation, feature-selection stability, calibration, clinical interpretability, and computational cost were also inconsistently reported. These findings highlight the need to evaluate RFE and MLP-based prediction beyond predictive accuracy alone. To address these challenges, this study delivers concrete, actionable guidelines for prevention of leakage, fold-safe RFE execution, and clinically aligned evaluation practices.

Keywords

How to Cite this Article

Bakar, W. A. W. A., Man, M., Zamri, N., Sabri, I. A. A., Sulaiman, N. S., Yacob, A., Aziz, N. S., Ahmad, S., & Josdi, N. L. N. (2026). Feature Selection for Machine Learning and Deep Learning-Based Heart Disease Prediction: A Systematic Review of Predictive Performance and Evaluation Practices. International Journal of Advanced Computer Science and Applications, 17(10). https://doi.org/10.14569/IJACSA.2026.0171050

Bakar, Wan Aezwani Wan Abu, et al.. "Feature Selection for Machine Learning and Deep Learning-Based Heart Disease Prediction: A Systematic Review of Predictive Performance and Evaluation Practices." International Journal of Advanced Computer Science and Applications, vol. 17, no. 10, 2026, https://doi.org/10.14569/IJACSA.2026.0171050.

@article{Bakar2026,
  title     = {Feature Selection for Machine Learning and Deep Learning-Based Heart Disease Prediction: A Systematic Review of Predictive Performance and Evaluation Practices},
  journal   = {International Journal of Advanced Computer Science and Applications},
  volume    = {17},
  number    = {10},
  year      = {2026},
  publisher = {The Science and Information Organization},
  author    = {Wan Aezwani Wan Abu Bakar and Mustafa Man and Nurnadiah Zamri and Ily Amalina Ahmad Sabri and Noor Suhana Sulaiman and Azliza Yacob and Nor Sukinah Aziz and Sharif Ahmad and Nur Laila Najwa Josdi},
  doi       = {10.14569/IJACSA.2026.0171050},
  url       = {https://doi.org/10.14569/IJACSA.2026.0171050}
}

Open Access — licensed under a Creative Commons Attribution 4.0 International License. Unrestricted use, distribution, and reproduction in any medium, even commercially, as long as the original work is properly cited.

Call for Papers - Important Dates

Volume 17 No. 12, December 2026
Paper Submission Due November 25, 2026
Review Notification December 15, 2026
Publication Date December 31, 2026