TensorLearn
Back to Course
Reinforcement Learning: Agents
Module 9 of 11

9. Curriculum Learning

1. Start Simple

You don't teach a child Calculus before Algebra. Don't ask an RL agent to solve the Hardest level immediately.

2. The Strategy

  1. Train on Easy environments (Reward is dense).
  2. Slowly increase difficulty (Reward becomes sparse).
  3. Result: Faster convergence and better generalisation.

Mark as Completed

TensorLearn - AI Engineering for Professionals