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 Video

Updated: Jul 15, 2026

Evaluation of Commercial-Off-The-Shelf Wrist Wearables to Estimate Stress on Students
12:51

Evaluation of Commercial-Off-The-Shelf Wrist Wearables to Estimate Stress on Students

Published on: June 16, 2018

Identifying predictors of student depression through validated machine learning pipelines.

Jacob Washton1, Tracy Owens2, Antony Sierra3

  • 1Algorithmic Medicine Laboratory, Department of Osteopathic Manipulative Medicine, College of Osteopathic Medicine, New York Institute of Technology, Old Westbury, NY, United States.

Frontiers in Medicine
|July 14, 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

Nested Fluid-Structure Interaction Predictive Modeling of Fetal Brain Stress During Maternal Trauma.

Biology·2026
Same author

Prevalence and Clinical Patterns of Piriformis Syndrome Among Actively Competing and Retired Elite Hockey Players.

Sports (Basel, Switzerland)·2026
Same author

Editorial for "Predicting Breath Hold Task Compliance From Head Motion".

Journal of magnetic resonance imaging : JMRI·2025
Same author

The Effect of Data Leakage and Feature Selection on Machine Learning Performance for Early Parkinson's Disease Detection.

Bioengineering (Basel, Switzerland)·2025
Same author

Electrocardiogram Abnormality Detection Using Machine Learning on Summary Data and Biometric Features.

Diagnostics (Basel, Switzerland)·2025
Same author

Impact of Convulsive Maternal Seizures on Fetus Dynamics.

International journal for numerical methods in biomedical engineering·2024

Machine learning models for student depression show unreliable predictions due to poor learning dynamics. Only optimized models with healthy learning curves accurately identify key risk factors like suicidal thoughts and academic pressure.

Area of Science:

  • Machine Learning
  • Public Health
  • Psychiatry

Background:

  • Student depression is a significant public health issue with 10-30% prevalence.
  • Machine learning can predict depression risk and identify key predictors.
  • Model reliability hinges on learning dynamics, often overlooked when performance metrics seem favorable.

Purpose of the Study:

  • Evaluate machine learning algorithms for student depression prediction.
  • Assess the impact of learning dynamics on model reliability and feature importance.
  • Identify optimal model configurations for accurate depression risk assessment in students.

Main Methods:

  • Compared six baseline algorithms and two RUSBoost pipelines on a 27,901-record student depression dataset.
  • Generated learning curves using progressively larger training data subsets (10%-100%).
Keywords:
RUSBoostfeature importancehyperparameter optimizationlearning curvesmachine learningmental health predictionmodel validationstudent depression

More Related Videos

Artificial Intelligence-Based System for Detecting Attention Levels in Students
06:37

Artificial Intelligence-Based System for Detecting Attention Levels in Students

Published on: December 15, 2023

Related Experiment Videos

Last Updated: Jul 15, 2026

Evaluation of Commercial-Off-The-Shelf Wrist Wearables to Estimate Stress on Students
12:51

Evaluation of Commercial-Off-The-Shelf Wrist Wearables to Estimate Stress on Students

Published on: June 16, 2018

Artificial Intelligence-Based System for Detecting Attention Levels in Students
06:37

Artificial Intelligence-Based System for Detecting Attention Levels in Students

Published on: December 15, 2023

  • Assessed algorithm health via monotonic validation accuracy and convergent training-validation gaps; Pipeline B featured hyperparameter optimization.

    • Main Results:

    • Baseline models and Pipeline A showed pathological learning dynamics (overfitting, non-monotonicity, oscillation).
    • Pipeline B demonstrated healthy learning dynamics with monotonic accuracy increases and gap convergence.
    • Pipeline B identified suicidal ideation history (1.0), academic pressure (0.57), and financial stress (0.31) as top predictors.

    Conclusions:

    • Aggregate performance metrics are insufficient for reliable scientific inference in machine learning models.
    • Learning curve diagnostics are crucial before interpreting feature importance for scientific conclusions.
    • Targeted interventions for suicidal thoughts, academic pressure, and financial stress are recommended for student depression.