Focus: Basics of AI/ML training.
-
Guided Project: Train TM model (smile vs neutral) → connect to Scratch → robot reacts.
-
Student Extension: Add third class (surprised face / custom gesture).
-
STEM Crossover: Math/Statistics + Ethics → Data samples & bias.
-
Compare training with 3 vs 30 samples (sample size & accuracy).
-
Demo bias: Train only on one student’s smile → robot fails on others. Retrain with group smiles → success.
-
Core Skills: Training models, dataset quality/quantity, fairness in AI.