Introduction to Artificial Intelligence

Mitch Marcus

CIS391
Fall, 2011
Welcome to CIS 391
 Course home page: http://www.cis.upenn.edu/~cis391

 Textbook: S. Russell and P. Norvig Artificial Intelligence: A

Modern Approach Prentice Hall, 2009, Third Edition

 Lecturer: Mitch Marcus, mitch(@)cis.upenn.edu, Levine 503

 TA: Ryan Kennedy, Levine 471, kenry (@) cis.upenn.edu
 TA: Qiuye Zhao, Levine 514, qiuye (@) seas.upen.edu

 Prerequisites: CSE 121 (Good knowledge of programming,

data structures)

1/12/2011 CIS 391 - Spring 2011 2

Grading & Homework
 Grading:
• 40% Homeworks (Final project will count for 1/4 of that
amount)
• 30% Midterm 1
• 30% Midterm 2
 Homework:
• Homework will be due at 11:00 a.m. on specified dates
• Late homeworks will be penalized unless an extension has
been granted in advance :
—15% reduction if 0-24 hours late
—30% reduction if handed in at the following class
—45% reduction if handed in two classes later
—No credit if more than a week late

1/12/2011 CIS 391 - Spring 2011 3

Tell us about yourself:
 What department and year are you in?

 Have you has a class in

• Probability and/or Statistics
• Logic

 Have you written code in Python?

1/12/2011 CIS 391 - Spring 2011 4

“Computers who will proud of us”

1/12/2011 CIS 391 - Spring 2011 5

A REAL Accomplishment: DARPA Grand Challenge

1/12/2011 CIS 391 - Spring 2011 6

A REAL Accomplishment: DARPA Urban Challenge

1/12/2011 CIS 391 - Spring 2011 7

Some Real Accomplishments of AI
 DARPA Grand Challenge – 123 miles through the desert
 DARPA Urban Challenge – Autonomous Driving in Traffic!
 Deep Space 1 – Remote Agent Experiment
 Deep Blue defeated the reigning world chess champion
Garry Kasparov in 1997
 Proved a mathematical conjecture (Robbins conjecture)
unsolved for decades
 Logistics Planning for 1991 Gulf War

 Computer Algebra Systems (Maple, Mathematica)

 Credit Evaluation, Fraud Detection
 Internet Search, Spam Filtering
 Handwritten character recognition (checks, US mail)
 Speech Recognition (Dictation)

1/12/2011 CIS 391 - Spring 2011 8

 Thinking humanly Thinking rationally

Acting humanly Acting rationally

What is AI?
Views of AI fall into four categories:

Thinking humanly Thinking rationally

Acting humanly Acting rationally

We will focus on "acting rationally"

1/12/2011 CIS 391 - Spring 2011 9

 Thinking humanly Thinking rationally
Acting humanly: Acting humanly Acting rationally
Turing Test
 Turing (1950) "Computing machinery and intelligence":
“Can machines think?”  “Can machines behave intelligently?”

 Operational test for intelligent behavior: the Imitation Game

 Predicted that by 2000, a machine might have a 30% chance of

fooling a lay person for 5 minutes

 Anticipated most major arguments against AI

 Suggested major components of AI: knowledge, reasoning,

language understanding, learning

1/12/2011 CIS 391 - Spring 2011 10

 Thinking humanly Thinking rationally
Thinking humanly: Acting humanly Acting rationally
cognitive modeling
 1960s "cognitive revolution": information-processing
psychology

 Requires scientific theories of internal activities of the brain

 How to validate? Requires

1) Predicting and testing behavior of human subjects (top-down)
or
2) Direct identification from neurological data (bottom-up)

 Both approaches (roughly, Cognitive Science and Cognitive

Neuroscience) are now distinct from AI
 Caveat: ACT-R & SOAR communities do computational
modelling of high level mental functions

1/12/2011 CIS 391 - Spring 2011 11

 Thinking humanly Thinking rationally
Thinking rationally: Acting humanly Acting rationally

"laws of thought"
 Aristotle: what are correct arguments/thought
processes?

 Several Greek schools developed various forms of logic:

notation and rules of derivation for thoughts; may or may
not have proceeded to the idea of mechanization

 Direct line through mathematics and philosophy to

modern AI

 Problems:
1. Not all intelligent behavior is mediated by logical deliberation
2. What is the purpose of thinking? What thoughts should I have?
3. Ignores the hard problem of perception

1/12/2011 CIS 391 - Spring 2011 12

 Thinking humanly Thinking rationally
Acting rationally: Acting humanly Acting rationally
rational agents
 Rational behavior: doing the right thing

 The right thing: that which is expected to

maximize goal achievement, given the available
information

 Doesn't necessarily involve thinking – e.g.,

blinking reflex – but thinking should be in the
service of rational action

1/12/2011 CIS 391 - Spring 2011 13

 Thinking humanly Thinking rationally

Acting humanly Acting rationally

Rational agents
 An agent is an entity that perceives and acts

 This course is about effective programming techniques for

designing rational agents

 Abstractly, an agent is a function from percept histories to

actions:
[f: P*  A]

 For any given class of environments and tasks, we seek the

agent (or class of agents) with the best performance

 Caveat: computational limitations make perfect rationality

unachievable
 design best program for given machine resources

1/12/2011 CIS 391 - Spring 2011 14

Rational Agent skills can include:

Vision Processing
Planning
Robotics
Natural Language Understanding

Search Reasoning Machine Learning

Representation of the World

Symbols (Logic, Numbers)

1/12/2011 CIS 391 - Spring 2011 15

Alternate approaches to AI
 Symbolic representations of the world
• Relations between entities
—―Lyle’s bicycle is red‖
– (isa B3241 bicycle) (color B3231 red) (owns B3241 P119)
– (isa P119 person) (name P119 ―Lyle‖)
• Explicit logical models
• Logical inference, Search
• Chess, Sudoko, computer games, …
 Instance-based and statistical models
• Prediction by look-up or by weighted combinations
—P(y=bicycle) = c0 + c1 x1 +c2 x2 + c3 x3 + …
• Implicit probabilistic models
• Machine vision, document retrieval, speech recognition, …

1/12/2011 CIS 391 - Spring 2011 16

Course Overview – First Half
Module 0: Introduction
• Intelligent Agents
• Python Programming (Week long introduction)
Module 1: Search Strategies
• Uninformed & Informed Search
• Constraint Satisfaction (Sudoku Solver)
• Adversarial Search (Game Playing)
• Homework: Build a solver for any Sudoku problem
• Final Project: Build a bot that competes against other game
bots to play a simple board game (Kingdoms)

1/12/2011 CIS 391 - Spring 2011 17

Course Overview – Second Half
Module 2: Machine Learning and Natural Language
Processing
• Introduction to Probability and Bayesian Networks
• Naive Bayes (Spam Filtering)
• Hidden Markov Models & Part of Speech Tagging
• Decision Trees
• Perceptrons and Support Vector Machines
• Homework: Build a spam filter using Naïve Bayes machine
learning

1/12/2011 CIS 391 - Spring 2011 18

Course Overview – Second Half (cont.)

Module 3: Knowledge Representation and Logic

• Logical Agents (Automated Wumpus adventurer)
• Topics in Knowledge Representation
• Homework: Build an agent that plays Hunt the Wumpus
using propositional logic theorem proving

1/12/2011 CIS 391 - Spring 2011 19