CS480/580 Introduction to Artificial Intelligence
Shuiwang Ji
General information Instructor: Shuiwang Ji
· Office hours: Monday and Wednesday, 4:30PM-5:30PM, or by appointment
· Office location: E&CS 3204· E-mail: [email protected]· Research interests: machine learning, data
mining, computer vision, computational biology Course Homepage:
http://www.cs.odu.edu/~sji/classes/fall2010/AI/
2
Textbooks Main text: Artificial Intelligence:
A Modern Approach, 3rd Edition, by Stuart Russell and Peter Norvig
http://aima.cs.berkeley.edu/
LISP programming: Practical Common Lisp by Peter Seibel
http://www.gigamonkeys.com/book/ Other LISP books also work
3
Grading Graduates and undergraduates will be graded
separately Homework:
· Undergraduate (30%): there will be 4-5 homework· Graduate: (25%): there will be 3-4 homework
Term paper (5%): ONLY for graduate students Project (30%): there will be 2-3 projects Exam (40%):
· Exam 1: 10%· Exam 2: 10%· Final exam 20%
4All homework and projects are strictly individual
LISP programming
Projects will involve LISP programming Use Lisp-in-a-box (link from the class page)
Easy to install and use A free book on LISP is available Partial code will ONLY be provided in LISP Need LISP tutorial lecture?
5
Project grading
Partial code in LISP will be provided Students are asked to program the core algorithms Some example inputs will be given, and the outputs
and analysis are graded Grading criteria
· General structure: 10%· Project report: 30%· Correctness: 60%
6
Course overview
Intelligent agent architecture Problem-solving by searching Constraint satisfaction problems (CSP) Propositional logic First-order logic Probabilistic inference and Bayesian networks Planning and Markov decision processes (MDP) Machine learning
7
Class homepage
The class is temporally run through the homepage: http://www.cs.odu.edu/~sji/classes/fall2010/AI
Switch to blackboard later on
Not hardcopy handouts, check class homepage/blackboard regularly
8
What is AI?
9
Two central questions
10
Humanly or rationally
Thinking or acting
Rational: does the “right thing” given what is knows
Definitions of AI
11
Humanly or rationally
Thin
king
or a
ctin
g
Think humanly: cognitive science
12
Do we want a machine that beats humans in chess or a machine that thinks like humans while beating humans in chess?
DeepBlue supposedly DOESN’T think like humans
Think rationally: law of thought
13
1. Not easy to take informal knowledge and state it in the formal terms required by logical notation
2. Reasoning on real-world problems is computationally demanding
Acting humanly: The Turing test
14
Mechanical flight became possible only when people decided to stop emulating birds…
Natural language processingKnowledge representation
Automated reasoningMachine learning
Acting rationally: rational agent
15
Making correct inference is sometimes part of being a rational agentCorrect inference is not all of rationalityThere are ways of acting rationally that cannot be said to involve inference
The rational agent approach
16
Think humanly Think rationally
Acting humanly Acting rationally
Humanly or rationally
Thin
king
or a
ctin
g
Rational agent
17
AI prehistory
18
Why AI?
19
Only thing Microsoft & Google agrees
20
“If you invent a breakthrough in artificial intelligence, so machines can learn," Mr. Gates responded, "that is worth 10 Microsofts." (Quoted in NY Times, Monday March 3, 2004)
No. 1: AI at human level in 10-20 year time frame· Sergey Brin & Larry Page
(independently, when asked to name the top 5 areas needing research. Google Faculty Summit, July 2007)
ENIAC: The beginning of computing age (1946)
21
Three fundamental questions in our age
Origin of the Universe Origin of life Nature of intelligence
22
Along with molecular biology, AI is regularly cited as the “field I would most like to be in” by scientists in other disciplines
The age before AI
23
I propose to consider the question: “Can machines think?” --Alan Turing, 1950
1956: A new field is born
24
We propose that a 2 month, 10 man study of artificial intelligence be carried out during the summer of 1956 at Dartmouth College in Hanover, New Hampshire.
- Dartmouth AI Project Proposal; J. McCarthy et al.; Aug. 31, 1955.
1997: Deep blue
25
I could feel human-level intelligence across the room -Gary Kasparov, World Chess Champion (human)
vs.
2005: Cars drive themselves
26
Stanley and three other cars drive themselves over a 132 mile mountain road
2005: Robots play soccer
27
2006: AI Celebrates its Golden Jubilee…
1956: A new field is born We propose that a 2 month, 10
man study of artificial intelligence be carried out during the summer of 1956 at Dartmouth College in Hanover, New Hampshire.
- Dartmouth AI Project Proposal; J. McCarthy et al.; Aug. 31, 1955.
Visual object recognition
29
Sample images from the PASCAL Visual Object Classes Challenge
Aeroplanes Bicycles Birds Boats Buses Cars
Cats Trains Cows Chairs Dogs Horses
Action recognition
30
Sample video frames from the TRECVID video surveillance evaluation
Autonomous robot
31
Mars Exploration Rover Learning Applied to Ground Robots (LAGR)
Medical diagnosis
32
..and thankfully You step in
to take CS 480/580Welcome!
What we will do?
Major AI topics: problem-solving by search, constraint satisfaction problems, logic reasoning, probabilistic reasoning, planning, machine learning
Practical implementation, such as 8-puzzle
34
Next class
Intelligent agent architecture Read Chapter 2 of AIMA
35