1. Artificial Intelligence Overview John Paxton Montana State University August 14, 2003

2. Montana State University

3. A Brief Bio 1985 The Ohio State University, B.S. Computer Science 1987 The University of Michigan, M.S. Computer Science 1990 The University of Michigan, Ph.D. Artificial Intelligence 2003 Montana State University – Bozeman, Professor of Computer Science

4. Talk Outline What is AI? Foundations History Areas Search Knowledge Representation Agents

5. What is AI? Science Approach 1.Systems that think like humans 2.Systems that act like humans Engineering Approach 1.Systems that think rationally 2.Systems that act rationally

6. Acting Humanly Turing Test (1950)

7. Thinking Humanly Cognitive Modelling Approach General Problem Solver (Newell and Simon, 1961)

8. Thinking Rationally The laws-of-thought approach Syllogisms (Aristotle) It is difficult to code the knowledge and to reason with it efficiently.

9. Acting Rationally Rational Agent Approach. The agent acts to achieve the best (or near best) expected outcome.

10. Foundations Philosophy (e.g. Where does knowledge come from?) Mathematics (e.g. What are the formal rules to draw valid conclusions?) Economics (e.g. How should we make decisions to maximize payoff?) Neuroscience (e.g. How do brains process information?)

11. Foundations Psychology (e.g. How do humans and animals think and act?) Computer Engineering (e.g. How can we build an efficient computer?) Control Theory (e.g. How can artifacts operate under their own control?) Linguistics (e.g. How does language relate to thought?)

12. History 1943-1955 Gestation. McCulloch-Pitts, Hebb, Turing Test 1956. Dartmouth Conference. 1952-1969. Great Expectations. Logic Theorist, GPS, Checkers, Lisp, Microworlds (calculus) 1966-1973. Reality. Machine translation (spirit == vodka), chess, intractability, fundamental limitations (Perceptrons).

13. History 1969-1979. Knowledge-Based Systems. Dendral (infer molecular structure) 1980-present. Commercial Products. 1986-present. Return of neural networks. 1987-present. Science. Hidden Markov Models. Neural Networks. Bayesian Networks. 1995-present. Intelligent Agents.

14. Areas Agents Artificial Life Machine Discovery and Data Mining Expert Systems Fuzzy Logic Game Playing Genetic Algorithms

15. Areas Knowledge Representation Learning Neural Networks Natural Language Processing Planning Reasoning Robotics

16. Areas Search Speech Recognition and Synthesis Virtual Reality Computer Vision

17. Search Missionaries and Cannibals Problem MMM CCC

18. Search Missionaries and Cannibals Solution MMM CCC MM CC MCMC CMMM CC MMMCCC MMM C CC MCMC MM CC MM CC MCMC MMM C

19. Types of Search Blind Search –Breadth-First Search –Depth-First Search Informed Search –Best-First Search –A* Search

20. Breadth-First Search MMM CCC MMM CC C MMM C CC MM CC MCMC

21. Minimax Search Commonly used to determine which move to make in a 2 player, strategy game. Deep Junior (Ban, Bushinsky, Alterman), the reigning computer chess champion uses minimax. Minimax requires an evaluation function.

22. Minimax Example Nim 4 (my move) 3 2 1 (your move) 2 1 1 1 (my move) 1 (your move)

23. Chess Example * * * * 3 0 -5 4 10 2 maximizer minimizer

24. Knowledge Representation Semantic Nets Fuzzy Logic First Order Predicate Calculus

25. Semantic Nets bird robin magpie ostrich yes no is-a can-fly

26. Fuzzy Logic Shaquille O’Neal is tall 5’0 6’0 7’0 tall 1.0 0.0

27. Fuzzy Logic Karim is tall (0.6) and a good teacher (0.9) = 0.6 Karim is tall or a good teacher = 0.9. Karim is not tall = 1.0 – 0.6 = 0.4

28. First Order Predicate Calculus Every Saturday is a weekend.  x Saturday(x)  weekend(x) Some day is a week day.  x day(x)  weekday(x)

29. Agents AGENT ENVIRONMENT sensors actuators

30. Rationality Factors Performance Measure Prior Knowledge Performable Actions Agent’s Prior Percepts

31. Rational Agent For each possible sensor sequence, a rational agent should select an action that is expected to maximize its performance measure, given the evidence provided by the sensor sequence and whatever built-in knowledge the agent has.

32. Agent Terminology Omniscience: the outcome of its actions are known. Impossible! Learning: taking actions in order to perform better (e.g. robot vacuum cleaner) Autonomy: the agent relies on its own sensors rather than built-in knowledge

33. Environments Fully observable vs. partially observable Deterministic vs. stochastic Episodic (classification) vs. sequential (conversation) Static vs. dynamic Discrete (chess) vs. continuous (taxi- driving) Single agent vs. multi-agent.

34. Types of Agents Reflex Model-Based Goal-Based Utility-Based Learning Combinations of the above!

35. Questions?