Knowledge Graph — Coursera Notes › Academic disciplines › Computer Science / Information Technology › Artificial Intelligence › Machine Learning & Data

Q-learning

concept · part of Machine Learning & Data

A value-based reinforcement learning algorithm that learns Q-values for state-action pairs using the Bellman equation, effective for discrete action spaces.

Inside Q-learning (2)

• Bellman equation — An equation used in Q-learning to update Q-values by balancing immediate and future rewards.
• Epsilon-greedy strategy — An exploration strategy used in Q-learning that alternates between random actions and exploiting learned Q-values.

Connections

• Alternative to Policy gradients
• Uses Epsilon-greedy strategy
• Uses Bellman equation

This is the text view of an interactive 3D knowledge graph — open this page with JavaScript enabled to explore it visually.