Win, Lose, Learn Logo
In the module Win, Lose, Learn, students learn how an AI agent can teach itself to play a game. Students explore the concept of reinforcement learning, specifically, Q-learning. Groups of students use the RAISE Playground to design and implement a game that learns from its mistakes and gets better as it plays with you.
 
 

AI Topics

Technology Requirements

Learner Profile

Standards Alignment
  • How can AI learn to make decisions in complex environments?

  • Reinforcement learning

  • Q-learning: Q-tables, the learning process, reward functions
  • 11 - 14 years old

  • Some experience with block-based programming
TBD

This module isn’t ready to be shared yet, but you can find other lessons about AI designed by MIT RAISE at DayOfAI.org.