MIT 6.034: Artificial Intelligence

Patrick Winston's classic AI course. A historical treasure covering the foundations of artificial intelligence.

About This Course

MIT 6.034, taught by the late Patrick Henry Winston, is a classic survey of artificial intelligence techniques. Winston was one of the pioneering figures in AI research at MIT, and his lectures are renowned for their clarity, wit, and depth.

While some topics have evolved since these lectures were recorded, the foundational concepts remain essential. Winston's teaching style makes complex ideas accessible and memorable.

What You Will Learn

  • Search: Depth-first, breadth-first, hill climbing, beam search, A*
  • Constraint Satisfaction: Domain reduction, constraint propagation
  • Games: Minimax, alpha-beta pruning
  • Learning: Nearest neighbors, identification trees, neural nets
  • Neural Networks: Perceptrons, backpropagation, deep nets (historical perspective)
  • Genetic Algorithms: Evolution-inspired optimization
  • Support Vector Machines: Maximum margin classifiers, kernels
  • Boosting: Combining weak learners into strong classifiers
  • Representations: Semantic nets, frames, rules, logic

Prerequisites

Basic programming experience. Introductory calculus and linear algebra helpful but not required.

Course content belongs to MIT OpenCourseWare. Licensed under Creative Commons BY-NC-SA.

Up next

fast.ai: Practical Deep Learning for Coders

Jeremy Howard's top-down approach to deep learning. Build real models from day one, then learn the theory. Covers CNNs, NLP, tabular data, and collaborative filtering.