n-Gage: Predicting in-class Emotional, Behavioural and Cognitive Engagement in the Wild
summary
n-Gage uses wearables (EDA, heart rate, motion) and room sensors (CO2, temperature, humidity) plus daily surveys to predict emotional, behavioural and cognitive engagement. A field deployment collects data from 23 students 144 classes 4 weeks and trains ML regressors with new EDA synchrony feature and environmental statistics. Results report average MAE 0.788 RMSE 0.975 across dimensions, show subject and time effects (higher cognitive scores in math, post‑lunch lows), and identify sensor combinations that most influence each engagement dimension.
AI-written summary · about