Plate IIMachine learning · 2026
Smartglasses Attention Modelling
Predicts which brands hold a smartglasses wearer's attention, and for how long, from 17,423 simulated detection events across 1,000 sessions.
- Store
- SQLite; detections under 0.70 confidence dropped
- Models
- Linear and logistic regression, decision tree, random forest; 5-fold cross-validation; k-means
- Stack
- Python, scikit-learn, pandas, NumPy, SQLite, matplotlib
Dwell time by brand, sTop 5 of 43
Simulated sessions · n = 17,423 · θ ≥ 0.70