Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

A comparative and interpretable machine learning framework for reliable diabetes risk prediction.

Talha Farooq Khan1, Mariyam Saeed1, Majid Hussain1

  • 1Department of Computer Science, The University of Faisalabad, Faisalabad, 38000, Pakistan.

Scientific Reports
|July 7, 2026
PubMed
Summary

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

A lab-scale prototype development of Lignin-Formaldehyde (LF) resin: A bio-renewable adhesive for wood panel industries.

PloS one·2026
Same author

Effect of substituting fishmeal with black soldier fly and mealworm larvae meal on nutrient digestibility, body and mineral composition of <i>Labeo rohita</i>.

Open veterinary journal·2026
Same author

IoMT-Blockchain framework for secure and real-time heart disease monitoring using hybrid black-winged optimized spherical structural graph convolutional neural networks.

Scientific reports·2026
Same author

A graph-based evaluation framework for smart disaster response systems using IoT design quality metrics.

Scientific reports·2026
Same author

Tensor enhanced chest cancer classification via CNN and Vision Transformer models.

PloS one·2026
Same author

Self-Supervision-Enabled Compounded Multi-Modal Feature-Learning Network for Classifying Depressive States with Fine-Grained Emotions Using Wearable Sensors.

Biosensors·2026

This study introduces an explainable machine learning model for accurate diabetes prediction, achieving 94% accuracy. It overcomes data imbalance and interpretability issues for reliable healthcare decision support.

Area of Science:

  • Medical Informatics
  • Machine Learning in Healthcare
  • Computational Statistics

Background:

  • Diabetes mellitus is a prevalent metabolic disease with severe health consequences.
  • Early and accurate diagnosis is crucial for effective diabetes management and improved patient outcomes.
  • Existing machine learning models for diabetes prediction often suffer from imbalanced datasets, limited interpretability, and suboptimal validation, hindering clinical usability.

Purpose of the Study:

  • To develop and validate a powerful and explainable machine learning model for predicting diabetes.
  • To address challenges of class imbalance and enhance model interpretability in diabetes prediction.
  • To provide a clinically credible and reliable machine learning solution for diabetes risk assessment.

Main Methods:

Keywords:
Class imbalanceDiabetes predictionExplainable AI (SHAP)Machine learningMedical data analysisModel interpretabilityStratified cross-validation

Related Experiment Videos

  • Utilized the Pima Indians Diabetes Dataset with eight clinical variables.
  • Employed the Synthetic Minority Over-sampling Technique (SMOTE) to balance the dataset.
  • Implemented stratified 10-fold cross-validation for robust model evaluation.
  • Trained and tested four classifiers: Logistic Regression, Naive Bayes, AdaBoost, and XG Boost.
  • Incorporated logistic regression coefficients and SHAP values for model interpretability.

Main Results:

  • Achieved an overall accuracy of approximately 94% on an unseen test set.
  • Demonstrated strong precision and recall for the minority diabetic class.
  • The proposed framework ensured leakage-free validation and robust cross-validation.
  • Identified critical clinical features influencing diabetes prediction through explainable AI techniques.

Conclusions:

  • The combination of class balancing, cross-validation, and explainable machine learning yields reliable and clinically credible diabetes predictions.
  • The developed framework offers a robust and interpretable solution for healthcare decision support in diabetes risk assessment.
  • Rigorous methodology and interpretability are essential for developing trustworthy machine learning applications in healthcare.