9-12:06:Model Training, Evaluation, and Performance Metrics

Not all accuracy is equal. Students train a real model, read a confusion matrix, and learn precision, recall and F1 – then see how the numbers can reveal bias.

“My model is 95% accurate” can hide a model that never catches the rare, important case. Learning to read a confusion matrix and choose the right metric is the difference between a demo and a trustworthy system.

What students do
• Not All Accuracy is Equal: a medical-diagnosis scenario where 95% accuracy isn’t good enough
• Let’s Train a Real Model – a classification model in Google Colab
• Meet the Confusion Matrix, then Precision, Recall and F1 – what each one tells you
• When Metrics Reveal Bias, Ethical Reflections, and Wrapping Up – You’re Now a Model Analyst

By the end, students can
• Evaluate a model properly
• Choose the right metric for the problem
• Connect metrics to fairness

Learning focus (Fink’s Taxonomy of Significant Learning): Application · Learning How to Learn · Caring.

Hands-on coding in Google Colab. Part of Grades 9–12: AI & You, Module 2 – Advanced AI/ML & Generative AI with Python, Week 2. Next lesson: Advanced OOP for Robust AI Applications & Project Showcase.