Lesson Plan

• Describe the core concepts of Machine Learning, Deep Learning and Reinforcement Learning.
• Explain why each method is chosen for particular problem types.
• Compare the strengths and limitations of ML, DL and RL.
• Apply a simple example to illustrate policy learning in RL.
• Evaluate how hardware advances enable deep‑learning applications.

• Projector and screen
• Laptop with Python / Jupyter Notebook
• Pre‑written notebook demonstrating a neural network and a Q‑learning agent
• Printed handout summarising key algorithms and the comparison table
• Whiteboard and markers
• Internet access for short video demo

Begin with a 1‑minute video of a self‑driving car navigating a city to spark curiosity. Review that students already know basic programming and supervised learning. State that by the end of the lesson they will be able to explain when to use ML, DL or RL and justify their choices.

1. Do‑now (5'): Quick digital quiz on definitions of ML, DL and RL.
2. Mini‑lecture (10'): Concise overview of each method, key algorithms and typical data.
3. Interactive demo (15'): Run the Jupyter notebook – first a simple CNN classifying images, then a GridWorld Q‑learning agent.
4. Group activity (10'): In small groups, analyse three real‑world scenarios and decide which AI approach is most suitable; fill in a comparison chart.
5. Guided practice (10'): Students tweak hyper‑parameters in the notebook and observe performance changes.
6. Check for understanding (5'): Exit ticket – write one sentence explaining why RL is ideal for game‑playing agents.

Summarise the three approaches, highlighting their appropriate use‑cases and the role of modern hardware. Collect exit tickets and remind students of the homework: read the assigned AI‑ethics chapter and write a brief reflection on the societal impact of deep learning.