DM828 - Introduction to Artificial Intelligence

The course covers basic elements from different areas of Artificial Intelligence.

All reading references are from the text book: S. Russell and P. Norvig. "Artificial Intelligence: A Modern Approach". Third Edition. Prentice Hall, 2010

Bold indicates topics that must be known well

1. History and Philosophical aspects of AI
   • Lecture 1 (Class 1)
     • Introduction
     • What is AI, History of AI, Application Examples [1.1, 1.3, 1.4; pp 1-33]
     • Intelligent Agents [pp. 34-63]
       • Enviroments and their properties [2.1, 2.3]
       • Rationality [2.2]
       • Perception action cycle and AI programs [2.4]
2. Problem solving by searching [pp. 64-119]
   • Lecture 2 (Class 2)
     • Problem Solving Agents [3.1, 3.3]
     • Examples [3.2]
     • Uninformed search, Breadth-first search, Uniform-cost search, Depth-first search, Depth-limited search, Iterative deepening search, Bidirectional Search [3.4]
     • Informed (Heuristic) Search [3.5, 3.6]
       • greedy search
       • Astar search
       • Memory bounded Astar
3. Uncertain knowledge and probabilistic reasoning
   • Lecture 3 (Class 4)
     • Probability and probability calculus [13.1-13.5, pp 480-499]
     • The Wumpus world probabilistic reasoning example [7.2, 7.3, 13.6 pp 236-243, 499-502]
   • Lecture 4 (Class 5)
     • Bayesian Networks, Semantics, Hybrid variables [14.1-14.3, pp 510-522]
   • Lecture 5 (Class 7)
     • Exact inference by enumeration and by variable elimination [14.4]. Complexity.
     • Stochastic inference [14.5]
       • prior sampling
       • rejection sampling
       • likelihood weighting
       • Markov Chain Monte Carlo
   • Lecture 6 (Class 8)
     • Graphical models for sequential data
       • Markov processes [15.1]
       • Inference: filtering and prediction, smoothing, most likely explanation [15.2]
       • Hidden Markov models [14.3]
       • Kalman filters [14.4]
       • Dynamic Bayesian Networks and Particle Filtering [15.5]
4. Machine learning
   • Supervised Learning [18.2]
   • Lecture 7 (Class 10)
     • Decision Trees [18.3]
     • k-Nearest Neighbor [18.8]
     • Linear Models [18.6]
     • Non-parametric regression [18.8]
   • Lecture 8 (Class 11)
     • Artificial Neural Networks [18.7]
     • Other issues: evaluation [18.5], learning theory [18.6], ensemble methods [18.10]
   • Lecture 9 (Class 13)
     • Learning parameters in Graphical Models [20.2]
       • Bayesian learning (small data set) [20.1, 20.2, 20.2.4]
       • Maximum likelihood learning (large data set) [20.1, 20.2.1, 20.2.2, 20.2.3]
       • Naive Bayes' Networks
   • Unsupervised Learning
     • k-means algorithm [see wikipedia]
     • Expectation Maximization algorithm [20.3.1]
   • Lecture 10 (Class 14)
     • Markov Decision Processes [17.1]
       • Value iteration algorithm [17.2.1, 17.2.2]
   • Reinforcement learning [21.1]
     • Passive RL [21.2.1], Termporal-difference learning [21.2.3, Sutton & Barto, 6.1 ]
     • Active RL, Q-learning [21.3]
5. Games and Adversarial Search
   • Minimax algorithm [5.2]
   • Alpha-beta pruning [5.3]
   • Cutoffs and evaluation functions [5.3]
   • Expectimax and Expectiminimax [5.5]
6. Knowledge representation and logic reasoning
   • Propositional logic [ch. 6]
   • First Order Logic [ch. 8]
   • Inference in propositional logic [ch. 5]
7. Applications:
   • Natural Language Processing [22.1]
     • Spam classifier [22.2]
     • Speech recognition [23.5]
     • Machine Translation
       • Statistical machine translation [23.4]
       • Rule based machine translation (grammars) [23.1-23.3.1]