What are the principal types of machine learning for AI agents?

The principal types of machine learning are broadly categorized based on the nature of the learning signal or feedback available to the learning system. The three classical paradigms are Supervised Learning (learning from labeled data), Unsupervised Learning (finding patterns in unlabeled data), and Reinforcement Learning (learning via trial-and-error with a reward system). Due to its transformative impact and unique architecture using multi-layered neural networks, Deep Learning is now widely considered a fourth principal type or a major subfield that spans the other three. This classification is standard in modern computer science and AI curricula.

A. Dynamic programming and Q-learning are specific algorithms and techniques used within reinforcement learning, not principal types.

B. The Markov decision process and the Bellman equation are mathematical frameworks foundational to reinforcement learning, not top-level categories.

C. Clustering and association rules are specific tasks or methods within the broader category of unsupervised learning.

1. Ng, A. (n.d.). CS229 Machine Learning Course Materials. Stanford University. The course syllabus and lecture notes define supervised, unsupervised, and reinforcement learning as the primary paradigms of machine learning.

2. Goodfellow, I., Bengio, Y., & Courville, A. (2016). Deep Learning. MIT Press. Chapter 5, "Machine Learning Basics," pp. 96-98. This foundational text introduces supervised and unsupervised learning, with later sections dedicated to deep learning as a modern approach.

3. Sutton, R. S., & Barto, A. G. (2018). Reinforcement Learning: An Introduction. MIT Press. This book is the standard academic text defining reinforcement learning as a major branch of machine learning.