Download ppt

Survey
yes no Was this document useful for you?
   Thank you for your participation!

* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project

Document related concepts

Technological singularity wikipedia , lookup

Artificial intelligence in video games wikipedia , lookup

Philosophy of artificial intelligence wikipedia , lookup

History of artificial intelligence wikipedia , lookup

Embodied cognitive science wikipedia , lookup

Intelligence explosion wikipedia , lookup

Ethics of artificial intelligence wikipedia , lookup

Existential risk from artificial general intelligence wikipedia , lookup

Transcript
Please pick up a
copy of the course
syllabus from the
front desk.
http://www.pami.uwaterloo.ca/~khoury/ece457
ECE457 Applied Artificial Intelligence
R. Khoury (2007)
Page 1
Introduction to AI
ECE457 Applied Artificial Intelligence
Fall 2007
Lecture #1
Outline

What is an AI?



Russell & Norvig, chapter 1
Agents
Environments

Russell & Norvig, chapter 2
ECE457 Applied Artificial Intelligence
R. Khoury (2007)
Page 3
Artificial Intelligence

Artificial intelligence is all around us


Computer players in
video games
Robotics


Expert systems


Assembly-line robots,
auto-pilot, Mars
exploration robots,
RoboCup, etc.
Medical diagnostics,
business advice,
technical help, etc.
Natural language

Spam filtering,
translation, document
summarization, etc.
ECE457 Applied Artificial Intelligence
R. Khoury (2007)
Page 4
What is an AI?

Systems that…
Humanly
Neural
Think
networks
ELIZA



Rationally
Theorem
proving
Deep Blue
Act
Rationality vs. Humans: emotions, instincts,
etc.
Thinking vs. acting: Turing test vs. Searle’s
Chinese room
Engineers (and this course) focus mostly on
rational systems
ECE457 Applied Artificial Intelligence
R. Khoury (2007)
Page 5
Act Rationally


Perceive the environment, and act so as to
achieve one’s goal
Not necessary to do the best action






There’s not always an absolutely best action
There’s not always time to find the best action
An action that’s good enough can be acceptable
Example: Game playing
Sample approach: Tree-searching strategies
Problem: Choosing what to do given the
constraints
ECE457 Applied Artificial Intelligence
R. Khoury (2007)
Page 6
Think Rationally




Uses logic to reach a decision or goal
via logical inferences
Example: Theorem proving
Sample approach: First-order logic
Problems:



Informal knowledge
Uncertainty
Search space
ECE457 Applied Artificial Intelligence
R. Khoury (2007)
Page 7
Think Rationally
1.
2.
3.
X = Y/Z  XZ = Y
X=Y
X+Z=Y+Z
X*Y+X*Z
X * (Y + Z)
a.
b.
c.
d.
e.
4.
5.
6.
b² = AH * c
a² = BH * c
a² + b² =
BH * c + AH * c
a² + b² =
c * (AH + BH)
a² + b² = c²
b/c = AH/b
a/c = BH/a
AH + BH = c
ECE457 Applied Artificial Intelligence
R. Khoury (2007)
Page 8
Act Humanly



“Turing-test” AI
Improve human-machine interactions
up to human-human level
Drawbacks:


In some cases, requires dumbing down the
AI
Lots of man-made devices work well
because they don’t imitate nature
ECE457 Applied Artificial Intelligence
R. Khoury (2007)
Page 9
Think Humanly



Cognitive science
Neural networks
Helps in other fields


Computer vision
Natural language processing
ECE457 Applied Artificial Intelligence
R. Khoury (2007)
Page 10
Rational Agents

An agent has

A rational agent has
an agent program
that allows it to do
the right action given
its precepts
ECE457 Applied Artificial Intelligence
Sensors
Actions

Percepts

Sensors to perceive
its environment
Actuators to act upon
its environment
Environment
Actuators
Agent
Program
R. Khoury (2007)
Page 11
Types of Agents

Simple reflex agent


Model-based agent


Considers what will happen given its actions
Utility-based agent


Keeps track of perception history
Goal-based agent


Selects action based only on current perception of
the environment
Adds the ability to choose between
conflicting/uncertain goals
Learning agent

Adds the ability to learn from its experiences
ECE457 Applied Artificial Intelligence
R. Khoury (2007)
Page 12
Simple Reflex Agent
Environment
Percepts
Actions
Actuators
Sensors
Current
State
Selected
Action
If-then
Rules
ECE457 Applied Artificial Intelligence
R. Khoury (2007)
Page 13
Simple Reflex Agent


Dune II (1992) units were
simple reflex agents
Harvester rules:





ECE457 Applied Artificial Intelligence
IF at refinery AND not empty
THEN empty
IF at refinery AND empty
THEN go harvest
IF harvesting AND not full
THEN continue harvesting
IF harvesting AND full
THEN go to refinery
IF under attack by infantry
THEN squash them
R. Khoury (2007)
Page 14
Model-Based Agent
Environment
Percepts
Actions
Actuators
Sensors
Current
State
Previous
perceptions
Selected
Action
World changes
Impact of actions
ECE457 Applied Artificial Intelligence
R. Khoury (2007)
If-then
Rules
Page 15
Goal-Based Agent
Environment
Percepts
Actions
Actuators
Sensors
Current
State
Previous
perceptions
State if I do
action X
Selected
Action
World changes
Goal
Impact of actions
ECE457 Applied Artificial Intelligence
R. Khoury (2007)
Page 16
Utility-Based Agent
Environment
Percepts
Actions
Actuators
Sensors
Current
State
Previous
perceptions
State if I do
action X
Happiness in
that state
World changes
Selected
Action
Utility
Impact of actions
ECE457 Applied Artificial Intelligence
R. Khoury (2007)
Page 17
Learning Agent
Environment
Percepts
Actions
Actuators
Sensors
Performance Element
Critic
Feedback
Knowledge
Changes
Learning
Problem
Element Learning Goals Generator
ECE457 Applied Artificial Intelligence
Performance
standard
R. Khoury (2007)
Page 18
Properties of the Environment

Fully observable vs. partially observable



Deterministic vs. stochastic vs. strategic



See everything vs. hidden information
Chess vs. Stratego
Controlled by agent vs. randomness vs.
multiagents
Sudoku vs. Yahtzee vs. chess
Episodic vs. sequential


Independent episodes vs. series of events
Face recognition vs. chess
ECE457 Applied Artificial Intelligence
R. Khoury (2007)
Page 19
Properties of the Environment

Static vs. dynamic vs. semi-dynamic



Discrete vs. continuous



World waits for agent vs. world goes on without
agent vs. world waits but agent timed
Translation vs. driving vs. chess with timer
Finite distinct states vs. uninterrupted sequence
Chess vs. driving
Single agent vs. cooperative vs. competitive


Alone vs. team-mates vs. opponents
Sudoku vs. sport team vs. chess
ECE457 Applied Artificial Intelligence
R. Khoury (2007)
Page 20
Properties of the Environment

Crossword Puzzle


Monopoly


Fully observable, stochastic, sequential, static,
discrete, competitive multi-agent
Driving a car


Fully observable, deterministic, sequential, static,
discrete, single-agent
Partially observable, stochastic, sequential,
dynamic, continuous, cooperative multi-agent
Assembly-line inspection robot

Fully observable, deterministic, episodic, dynamic,
continuous, single-agent
ECE457 Applied Artificial Intelligence
R. Khoury (2007)
Page 21