ChatWhole Learn
← Back to Data Science

All Topics

Advertisement

Learn/Data Science/Machine Learning

Temporal Difference Learning

Topic: RL

Advertisement

TD Learning Fundamentals

Learn from bootstrapping.

TD(0)

One-step bootstrap. Update towards TD target.

TD(λ)

Multi-step returns. Eligibility traces.

Eligibility Traces

Backward view of learning. λ parameter.

Key Takeaways

  1. TD learning bootstraps
  2. TD(λ) with eligibility traces
  3. Multi-step returns

Advertisement

← Policy Gradient MethodsDyna Architecture →

Advertisement

Need More Practice?

Get personalized data science help from ChatWhole's AI-powered platform.

Get Expert Help →