A General Introduction to
Artificial Intelligence
Course Information
Course Instructor: Dr Nguyen Xuan Hoai
Working Office: Room 311-3, Building 302.
E-mail: nxhoai@gmail.com
Phone: 8801611
Course Website:
Course Prerequisites:
Computer Algorithms and Data Structures
prerequisites, basic knowledge in Computer Science
Math (discrete structures, basic calculus and
probability) is assumed.
Required Skills:
Working knowledge of programming languages (C++
or JAVA is preferred).
Course Information
Course Description:
The course introduces the essential concepts and issues in artificial
intelligence. Topics include intelligent problem solving with search,
knowledge representation and inference, intelligent agents,
intelligent planning, and machine learning.
Course Textbooks:
Stuart Russell and Peter Norvig, Artificial Intelligence: A Modern
Approach, 2nd Edition, Prentice Hall, 2003.
Bigus and J. Bigus, Intelligent Agent Programming with JAVA, 2nd
Edition, John Wiley & Sons, 2001.
M. Ginsberg, Essentials of Artificial Intelligence, Morgan Kaufmann,
1993.
E. Rich & K. Knight, Artificial Intelligence, McGraw-Hill, 1991.
Course Grading:
10% Programming Assignments.
48% Course Projects (4).
42% Course open book Exams (In Class, 21% Midterm, 21% Final).
Contents
What
is Artificial Intelligence (AI)?
 AI related areas.
 Brief history of AI.
 Applications of AI.
 Core issues in AI
 What will be in the course?.
What is AI?


Definitions of AI have been somewhat controversial
(because of A and because of I).
Two main school of thoughts on what AI is: Strong
AI and Weak AI.
(see “The Artificial Minds”- MIT Press 1995 by
Franklins)
Strong AI
Strong AI implication: Intelligent agents can become
sapients (human-being, self-aware).
(AI researchers in the early age
and....Hollywood!!!)
Weak AI
Weak AI implication: Intelligent agents could only
simulate human-being some human being behaviors
(Widely accepted now)
Views on AI
Views on AI fall into 4 categories
Thinking Humanly
Thinking Rationally
Acting Humanly
Acting Rationally
The view of the course: acting rationally
Acting Humanly
Subjected to study the human intelligence.

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.

Acting Humanly
Two main approaches:
Top-down: Cognitive science  Symbolism.
Bottom-up: Neural and Brain Science 
Connectionism.
Acting Humanly
Turing Test (1950):
Predicted that by 2000, a machine might have a 30% chance of
fooling a lay person for 5 minutes.
Anticipated all major arguments against AI in following 50 years
Suggested major components of AI: knowledge, reasoning,
language understanding, learning.
Thinking rationally:
"laws of thought"
What are the rules (laws) of thought?
Aristole  Gorge Bool  David Hilbert = Logic
Thinking 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 (logic-based agents).
Problems:




1.
2.
Not all intelligent behavior is mediated by logic.
What is the purpose of thinking? What thoughts should I have?
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.
Rational Agents





An agent is an entity that perceives and acts
This course is about 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.
Rational Agents
Advantages of the view:
- Intelligence does not necessary require thinking
and/or reasoning.
- Intelligence is not necessary attach to human or
living creatures.


Intelligence can be in a process.
Intelligence can be obtained by cooperation of a
swarm of agents.
Rational Agents
Examples:
Evolutionary Intelligence, Swarm Intelligence.
Some Definitions of AI
from AI Books


"The exciting new effort to make computer think
… machine with minds, in the full and literal
sense" (Haugeland, 1985).
"Activities that we associate with human thinking,
activities , as such decision-making, problem
solving, learning" (Bellman, 1978).
Some Definitions of AI
from AI Books


"The art of creating machines that perform
functions that require intelligence when
performed by people" (Kurzweil, 1990).
"The study of how to make computers do
things, at the moment, people are better" (Rich
and Knight, 1991).
Some Definitions of AI
from AI Books


"The study of mental faculties through the use
of computational models" (Charniak and
McDermott, 1985).
"The study of the computations that make it
possible to perceive, reason, and act" (Winston,
1992).
Some Definitions of AI
from AI Books


"Computational Intelligence is the study of the
design of intelligent agents" (Poole et al., 1998).
"AI …. is concerned with intelligent behavior in
artifacts" (Nilsson, 1998) .
AI-Related Areas











Philosophy.
Cognitive science.
Neuroscience and Brain Theory.
Cybernetics and control theory.
Mathematical Logic.
Evolutionary Biology.
Social Intelligence.
Swarm Behavior.
Organization Theory.
Statistics.
.......
AI History
Three stages:
Symbolism (70-80) (Automated Reasoning and Proofing,
Expert Systems, Logic Programming,...).
Connectionism (80s-90s) (Neural Networks, Statistical
Learning, Support Vector Machines, Probabilistic Graph
Learning,....).
Evolutionary Computation (90s-?) (Evolutionary
Programming, Evolutionary Strategies, Genetic
Algorithms) , Intelligent Multi Agent Systems.
Abridged History of AI
1943 McCulloch & Pitts: Mô hình boole cho não bộ.
1950 Turing's "Computing Machinery and
Intelligence"
1956 Dartmouth meeting: "Artificial Intelligence “
was coined (Minsky?).
1956 Rosenblatt, Widrow and Hoff - PERCEPTRON
1950s Samuel's checker program,
Newell & Simon's Logic Theorist,
Gelernter's Geometry Engine.
1964 Evolutionary Strategies (Rechenberg et al.).
1964 Evolutionary Programming (L. Fogel).
1965 Robinson's complete algorithm for logical
reasoning.
Abridged History of AI
1969
Minsky and Papert - "PERCEPTRON"
1969-79 Knowledge-based systems (Expert and
Planning Systems) - Symbolism dominant time.
1980-85 AI became an industry.
1986: Rumelhart, Hinton, Williams - Back
Propangation learning algorithm for multi-layer
PERCEPTRON - the rebirth of neural networks.
1987 AI became an science.
1986-1995 Neural Networks, Machine Learning,
Approximate Reasoning, Fuzzy Systems,...
Connectionism time.
1995 - Evolutionary Computation, Natural
Computation, Intelligent Multi-Agent Systems.
Areas/Applications in AI













Natural Language Processing.
Automated Reasoning.
Knowledge-Based Systems.
Pattern Recognition.
Computer Vision.
Speech Processing.
Data Mining and Knowledge Discovery.
Intelligent Planning.
Intelligent Computer Games.
Multi-agent Systems.
Evolutionary and Natural Computation.
Artificial Life.
........
State of The Art









Deep Blue defeated the reigning world chess champion
Garry Kasparov in 1997
MYCIN (1984, Standford).
Proved a mathematical conjecture (Robbins conjecture)
unsolved for decades.
During the 1991 Gulf War, US forces deployed an AI
logistics planning and scheduling program that involved
up to 50,000 vehicles, cargo, and people
Gulf War 2 (2003), Artificial War.
NASA's on-board autonomous planning program
controlled the scheduling of operations for a spacecraft.
New washing machine generation using NeuroFuzzy
Technology.
Human identification through eyes detection and
analysis at Heathrow airport using evolutionary
computation technique.
........
Core Issue in AI




Representation.
Reasoning.
Learning.
Interaction.
Course Details
Week 1: Introduction & Agents
Week 2: Problem Solving and Search.
Week 3: Informed Search.
Week 4: Modern Meta-Heuristic Search.
Week 5: CSP and Adversarial Search.
Week 6: Logic Agents.
Week 7: First Order Logic.
Week 8: Knowledge Representation, Mid-term exam.
Week 9: Planning.
Week 10: Reasoning under uncertainty 1
Week 11: Reasoning under uncertainty 2
Week 12: Learning 1
Week 13: Learning 2
Week 14: Learning 3
Week 15: Summary and Final exam.
10% Programming Assignment: Due third week.
4 course projects: week 5, week 9, week 12, week 15.
Intelligent Agents
Outline





Agents and environments
Rationality
PEAS (Performance measure, Environment,
Actuators, Sensors)
Environment types
Agent types
Agents





An agent is anything that can be viewed as
perceiving its environment through sensors and
acting upon that environment through actuators
Human agent: eyes, ears, and other organs for
sensors; hands,
legs, mouth, and other body parts for actuators.
Robotic agent: cameras and infrared range
finders for sensors;
various motors for actuators.
Agents and Environments



The agent function maps from percept histories
to actions:
[f: P*  A]
The agent program runs on the physical
architecture to produce f
agent = architecture + program
Vacuum-Cleaner World


Percepts: location and contents, e.g., [A,Dirty]
Actions: Left, Right, Suck, NoOp
A Vacuum-Cleaner Agent
Percept sequence
Action
[A, Clean]
Right
[A,Dirty]
Suck
[B, Clean]
Left
[B, Dirty]
Suck
[A, Clean], [A, Clean]
Right
[A, Clean], [A, Dirty]
Suck
….
….
[A, Clean], [A, Clean], [A, Clean]
Right
[A, Clean], [A, Clean], [A, Dirty]
Suck
….
….
Rational Agents



An agent should strive to "do the right thing",
based on what it can perceive and the actions it
can perform. The right action is the one that will
cause the agent to be most successful.
Performance measure: An objective criterion for
success of an agent's behavior.
E.g., performance measure of a vacuum-cleaner
agent could be amount of dirt cleaned up, amount
of time taken, amount of electricity consumed,
amount of noise generated, etc.
Rational Agents

Rational Agent: For each possible percept
sequence, a rational agent should select an
action that is expected to maximize its
performance measure, given the evidence
provided by the percept sequence and whatever
built-in knowledge the agent has.
Rational Agents



Rationality is distinct from omniscience (allknowing with infinite knowledge)
Agents can perform actions in order to
modify future percepts so as to obtain
useful information (information gathering,
exploration)
An agent is autonomous if its behavior is
determined by its own experience (with
ability to learn and adapt)
PEAS



PEAS: Performance measure, Environment,
Actuators, Sensors
Must first specify the setting for intelligent agent
design.
Consider, e.g., the task of designing an
automated taxi driver:




Performance measure
Environment
Actuators
Sensors
PEAS


Must first specify the setting for intelligent agent
design:
Consider, e.g., the task of designing an automated
taxi driver:




Performance measure: Safe, fast, legal, comfortable
trip, maximize profits.
Environment: Roads, other traffic, pedestrians,
customers.
Actuators: Steering wheel, accelerator, brake, signal,
horn.
Sensors: Cameras, sonar, speedometer, GPS, odometer,
engine sensors, keyboard.
PEAS





Agent: Medical diagnosis system
Performance measure: Healthy patient, minimize
costs, lawsuits
Environment: Patient, hospital, staff
Actuators: Screen display (questions, tests,
diagnoses, treatments, referrals)
Sensors: Keyboard (entry of symptoms, findings,
patient's answers)
PEAS





Agent: Part-picking robot
Performance measure: Percentage of parts in
correct bins
Environment: Conveyor belt with parts, bins
Actuators: Jointed arm and hand
Sensors: Camera, joint angle sensors
PEAS





Agent: Interactive English tutor
Performance measure: Maximize student's score
on test
Environment: Set of students
Actuators: Screen display (exercises,
suggestions, corrections)
Sensors: Keyboard
Environment Types



Fully observable (vs. partially observable): An agent's
sensors give it access to the complete state of the
environment at each point in time.
Deterministic (vs. stochastic): The next state of the
environment is completely determined by the current state
and the action executed by the agent. (If the environment
is deterministic except for the actions of other agents, then
the environment is strategic).
Episodic (vs. sequential): The agent's experience is divided
into atomic "episodes" (each episode consists of the agent
perceiving and then performing a single action), and the
choice of action in each episode depends only on the
episode itself.
Environment Types



Static (vs. dynamic): The environment is
unchanged while an agent is deliberating. (The
environment is semidynamic if the environment
itself does not change with the passage of time
but the agent's performance score does)
Discrete (vs. continuous): A limited number of
distinct, clearly defined percepts and actions.
Single agent (vs. multiagent): An agent
operating by itself in an environment.
Environment Types
Fully observable
Deterministic
Episodic
Static
Discrete
Single agent


Chess with
a clock
Yes
Strategic
No
Semi
Yes
No
Chess without
a clock
Yes
Strategic
No
Yes
Yes
No
Taxi driving
No
No
No
No
No
No
The environment type largely determines the agent design
The real world is (of course) partially observable,
stochastic, sequential, dynamic, continuous, multi-agent
Agent Functions and Programs



An agent is completely specified by the agent
function mapping percept sequences to actions
One agent function (or a small equivalence class)
is rational
Aim: find a way to implement the rational agent
function concisely
Table-lookup Agent
Function TABLE-DRIVEN-AGENT(percept) return an action
Static: percepts, a sequence, initially empty.
table, a table of actions, indexed by percept sequences, initially
fully specified.
append percept to the end of percepts
action  LOOKUP(percepts, table)
return action

Drawbacks:
 Huge table
 Take a long time to build the table
 No autonomy
 Even with learning  long time to learn the table entries
Agent Program for
A Vacuum-Cleaner Agent
Function REFLEX-VACUUM-AGENT (location, status)
Rerurn an action
if status=Dirty then Suck
else if location=A then return Right
else if location=B then return Left
Agent Types
Four basic types in order of increasing generality:




Simple reflex agents
Model-based reflex agents
Goal-based agents
Utility-based agents
Simple Reflex Agents
Simple Reflex Agents
Function SIMPLE-REFLEX-AGENT (percept)
Return an action
Static: rules, a set of condition-action rules
state  INTERPRET-INPUT (percept)
rule  RULE-MATCH (state, rules)
action  RULE-ACTION [rule];
return action
Model-Based Reflex Agents
Model-based Reflex Agents
Function REFLEX-AGENT-WITH-STATE (percept)
Return an action
Static state, a description of the current world state
rules, a set of condition-action rules
action, the most recent action, initially none
state  UPDATE-STATE (state, action, percept)
rule  RULE-MATCH (state, rules)
action  RULE-ACTION [rule]
return action
Goal-based Agents
Utility-based Agents
Learning Agents
Further Reading



Main textbook - chapter 2.
Course textbook 2 - chapter 2.
M. Wooldride, An Introduction to Multi-agent
Systems (Chapter 2).