Related Experiment Video
Updated: May 6, 2026
Evaluation of Commercial-Off-The-Shelf Wrist Wearables to Estimate Stress on Students
Published on: June 16, 2018
Mental illness risk prediction in high school students using artificial neural network
Samuel Encarnação1, Paula Fortunato Vaz2, Filipe Vaz3
1Department of Physical Activity and Sport Sciences, Universidad Autónoma de Madrid (UAM), Ciudad Universitaria de Cantoblanco, 28049 Madrid, Spain; Department of Sport Sciences, Instituto Politécnico de Bragança (IPB), 5300-253 Bragança, Portugal; Live Well Research Centre for Active Living and Wellbeing, Instituto Politécnico de Bragança, Portugal; CI ISE, Instituto Superior de Ciências Educativas do Douro, 4560 547 Penafiel, Portugal.
Introduction:
The sustainable development goals of the United Nations 2030 agenda, goal number 3 - Good health and well-being- align with student mental health.
Objective:
To conduct an artificial neural network (ANN) to predict the students' self-reported mental health dimensions.
Methods:
A cross-sectional and observational study enrolling sociodemographic and health state data from 2050 university students aged (18-30 years).
Results:
The best algorithm's result was by predicting the students' depressive state with 97 % accuracy (weighted average = [precision = 0.79 %, recall = 0.79 %, F-1 score 0 0.79 %, cross-validation (73 %)]), while dimensions such overall mental health self-perception (validation accuracy = 60 %) and lack of interest in performing their activities of daily living [(ADLs), validation accuracy = 67 %], presented inferior predictions.
Conclusions:
The ANN best predicted the university students' depressive state (73 %).
