CSSE 470 Artificial Intelligence

1
CSSE 470Artificial Intelligence

• Instructor
• Omid Madani

2
Course Info

• Instructor Omid Madani
• Email madanio_at_seattleu.edu
• Who is he?
• Recent PhD from UW CS
• Interests AI, theory of computation,..
• Adjunct Seattle U Instructor
• Office and hours 411b after class

3
Course Material

• Text Artificial Intelligence A Modern Approach
• Stuart Russel and Peter Norvig
• Recommended
• Machine Learning, by Tom Mitchell
• Modern Information Retrieval, R. Baeza-Yates and
  B. Ribeiro-Neto
• Papers
• Class web http//classes.seattleu.edu/computer_sc
  ience/csse470/Madani/cs470.html

4
Required Work

• Weekly or Biweekly homeworks (25)
• Final Exam (30)
• Two programming projects
• Text categorization using ML (25)
• Possibly Satisfiability Solver (20)
• What programming language?

5
Topic Sequence

• 20 lectures, 2 hrs each
• First Half
• Introduction
• Search (weeks 1 and 2)
• Learning (weeks 1 thru 4)
• Second Half (weeks 4 thru the end)
• Representation and Reasoning
• Modeling and use of Uncertainty
• Other topics Planning, etc.

6
What is AI?

• What is intelligence?
• The ways humans think..
• The ways humans behave ..
• The ways rational/intelligent things think..
• -The ways rational/intelligent things behave
• AI is the science of understanding intelligence
  and the art of making intelligent things

7
What is rationality?

• Book
• A system is rational if it does the right
  thing.
• Right thing?
• Doing the thing that would make the agent be most
  successful or maximize a performance measure
• based on agents limited resources
• What it has perceived, what it knows, available
  actions, and computational resources

8
What does AI do?

• Automation of problem solving
• Learning
• Memory (Knowledge Representation)
• Reasoning
• Acting
• Study of mental faculty through computational
  models
• Making computers do what people do better now (or
  did better at some point!)

9
Products of AI

• Better understanding of the constituents of
  Intelligence
• And so many applications
• Vision (object recognition and tracking, e.g.
  faces)
• Voice/speech/language (dialogue, translation, IR)
• Game playing (chess, robotic soccer)
• Robotics (manufacturing, house cleaning)
• Pattern discovery, prediction (financial, DNA,
  data mining)
• Expert systems

10
Surpassing Humansat Various Tasks
chess
arithmetic
scheduling
The span of human intelligent behavior

• Once a problem is solved, it is solved much
  better than humans
• success in limited domains? Perhaps, not any
  more..

11
Allied Fields

• Computer Science (parent of?)
• Mathematics
• Logic, probability and statistics, optimization,
• Cognitive Science, Neuroscience, Psychology,
  probability
• Philosophy (and linguistics,..)
• Engineering (electrical, etc.) adaptation,
  control
• Economics (decision sciences, utility,..)

12
A Brief History

• 40-50s AI as a field is born neurons, games,
  logic
• 50-60sYouthful enthusiasm search, microworlds
• 60-70sKnowledge-Based systems
• 80sAI becomes an Industry, expert systems,
  neural nets are back
• 90s Many techniques and applications, e.g. use
  of probability, learning, computer beats human
  chess champ, AI matures
• 2050 a team of robots has beaten human soccer
  champs