11 Beyond Supervised Learning

This chapter goes beyond the supervised learning of Chapter 7. It covers learning richer representation and learning what to do; this enables learning to be combined with reasoning. First we consider unsupervised learning in which the classifications are not given in the training set. This is a special case of learning belief network, which is considered next. Finally, we consider reinforcement learning, in which an agent learns how to act while interacting with an environment.

• 11.1 Clustering
  • 11.1.1 Expectation Maximization
  • 11.1.2 k-Means
  • 11.1.3 EM for Soft Clustering
• 11.2 Learning Belief Networks
  • 11.2.1 Learning the Probabilities
  • 11.2.2 Unobserved Variables
  • 11.2.3 Missing Data
  • 11.2.4 Structure Learning
  • 11.2.5 General Case of Belief Network Learning
• 11.3 Reinforcement Learning
  • 11.3.1 Evolutionary Algorithms
  • 11.3.2 Temporal Differences
  • 11.3.3 Q-learning
  • 11.3.4 Exploration and Exploitation
  • 11.3.5 Evaluating Reinforcement Learning Algorithms
  • 11.3.6 On-Policy Learning
  • 11.3.7 Assigning Credit and Blame to Paths
  • 11.3.8 Model-Based Methods
  • 11.3.9 Reinforcement Learning with Features
• 11.4 Review
• 11.5 References and Further Reading
• 11.6 Exercises