Download as pptx, pdf, or txt
Download as pptx, pdf, or txt
You are on page 1of 34

‫بسم هللا الرحمن الرحيم‬

King Abdulaziz University


Faculty of Computing and information Technology
Computer Science Department

CPCS-335 Introduction to Artificial Intelligence

Introduction

Imtiaz Hussain Khan, PhD Abdullah Basuhail, PhD


ihkhan@kau.edu.sa abasuhail@kau.edu.sa
CPCS335, Introduction to Artificial Intelligence, CS, FCIT, KAU, 1441H/2020G
1
Readings/References
 Textbook
• Artificial Intelligence: A Modern Approach by Stuart Russell and Peter Norvig. 3rd
Edition.
 Additional References
• Lecture slides
• Web will also be useful references.

CPCS335, Introduction to Artificial Intelligence, CS, FCIT, KAU, 1441H/2020G


2
Course Outline
1. Artificial Intelligence Fundamentals
2. AI Search Strategies
3. Knowledge Representation
4. Reasoning and Expert Systems
5. Introduction to Machine Learning
6. Introduction to Natural Language Processing
7. Introduction to Computer Vision & Robotics

CPCS335, Introduction to Artificial Intelligence, CS, FCIT, KAU, 1441H/2020G


3
Course Learning Outcomes
1. Describe the key components of the AI field and its relation and role in
different majors.
2. Describe the characteristics of a given problem that an intelligent
system must solve.
3. Apply an appropriate uninformed/informed search strategy for a
problem.
4. Compare and contrast basic search issues with game playing issues.
5. Apply resolution to a set of logic statements to answer a query.

CPCS335, Introduction to Artificial Intelligence, CS, FCIT, KAU, 1441H/2020G


4
Course Learning Outcomes (Cont.)
6. Explain the different components of an expert system.
7. Recognize different applications in NLP including language corpora.
8. Identify techniques for information extraction, language translation,
and text classification.
9. Apply Python NLTK to solve different NLP tasks.
10. List the differences among the major types of machine learning.
11. Identify examples of classification tasks, including the available input
features and output to be predicted.

CPCS335, Introduction to Artificial Intelligence, CS, FCIT, KAU, 1441H/2020G


5
Course Learning Outcomes (Cont.)
12. Use Bayes’ theorem to determine the probability of a hypothesis given
evidence.
13. Apply different classifiers and clustering techniques available in
Python/Weka to solve a given classification/clustering problem.
14. Describe the notion of robotics and machine vision.

CPCS335, Introduction to Artificial Intelligence, CS, FCIT, KAU, 1441H/2020G


6
Assessment & Evaluation

# *Assessment task Week Due Assessment (%) Score

1 Exam 1 6th week 20


2 Exam 2 12th week 20
3 Graded Lab Work TBA 25
4 Final Exam 15th week 35

CPCS335, Introduction to Artificial Intelligence, CS, FCIT, KAU, 1441H/2020G


7
Today’s Plan
Definitions
AI Goals
AI Approaches
Turing Test

CPCS335, Introduction to Artificial Intelligence, CS, FCIT, KAU, 1441H/2020G


8
Artificial Intelligence in Our Life
 Robots can help nurses in hospitals
to deliver stuff to different rooms

 Drones can record our fancy


videos while we do outdoor
activities

CPCS335, Introduction to Artificial Intelligence, CS, FCIT, KAU, 1441H/2020G


9
Artificial Intelligence in Our Life
 Calorie Mama makes instant nutrition and
calorie estimates from your meals
(click here to watch video)

 Translate Cam AI translates photo texts and


documents into any language with one touch
(click here to watch video)

CPCS335, Introduction to Artificial Intelligence, CS, FCIT, KAU, 1441H/2020G


10
Artificial Intelligence in Our Life
 Virtual Personal Assistants
• Siri, Google Now: Where is the nearest Albaik restaurant?
 Video Games
• e.g. Far cry & Call of duty
 Smart Cars
• e.g. Google’s self driving car & Tesla’s autopilot
 Fraud Detection
• Credit card fraud detection
 (. . .)
CPCS335, Introduction to Artificial Intelligence, CS, FCIT, KAU, 1441H/2020G
11
Artificial Intelligence Agents
 An agent is anything that can be viewed as
• perceiving its environment through sensors and
• acting upon that environment through actuators
 An agent is a computer system capable of
flexible and autonomous action in some
environment
 A rational agent should attempt 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
CPCS335, Introduction to Artificial Intelligence, CS, FCIT, KAU, 1441H/2020G
12
Agent Example
 Vacuum-cleaner world
• Percept: Location (e.g. A or B) and content (e.g.
Dirty or clean)
• Actions: Move left, move right, suck, NoOp
• Function: if the current square is dirty then suck,
otherwise move to the other square or standby

CPCS335, Introduction to Artificial Intelligence, CS, FCIT, KAU, 1441H/2020G


13
What is Intelligence?
Computational part of the (human) ability to achieve goals in the world
 Intelligent behaviour involves:
• ability to act in complex environments
• ability to learn from experience
• ability to think and reason
• ability to perceive relations (in the world)
• ability to use tools
• (…)
 Varying kinds and degrees of intelligence occur in people, many animals
and even in some machines 14
CPCS335, Introduction to Artificial Intelligence, CS, FCIT, KAU, 1441H/2020G
What is Artificial Intelligence (AI)?
 The science and engineering of making intelligent machines (John
McCarthy)
 The study of ideas that make people intelligent and incorporate those
ideas into computers (Patrick Henry)
 The study of making computers do things which, at the moment, people
are better (Elaine Rich)
 Getting computers to do tasks which require human intelligence
(Anonymous)

CPCS335, Introduction to Artificial Intelligence, CS, FCIT, KAU, 1441H/2020G


15
What is Artificial Intelligence (AI)?
 AI is a multidisciplinary field
• Psychology: To investigate how human mind/brain works
• Philosophy: Theories of reasoning and learning
• Mathematics: Theories of logic, probability, decision making and computation
• Linguistics: The meaning and structure of language
• Computer Science: Makes AI a reality by developing computational models

CPCS335, Introduction to Artificial Intelligence, CS, FCIT, KAU, 1441H/2020G


16
Why Artificial Intelligence?
 Computers can solve some difficult problems much quicker than human
• Computing GCD of two numbers
• Solving complex integration problems
• Computing product of, say, four numbers
 Human can solve some simple problems much elegantly than computers
• Navigating in a busy street
• Recognizing voice of a person
 The first law of AI
• Easy problems are hard and hard problems are easy

CPCS335, Introduction to Artificial Intelligence, CS, FCIT, KAU, 1441H/2020G


17
Strong vs. Weak AI
 Strong AI refers to the AI research that aims to replicate human
intelligence completely
• Overall intellectual ability of machine is indistinguishable from human
• Machines can beat human in all respects
 Weak AI refers to the AI research that aims to apply partial human
intelligence in specific problem solving
• Chess playing program – Deep blue

CPCS335, Introduction to Artificial Intelligence, CS, FCIT, KAU, 1441H/2020G


18
Engineering-based AI Goal
 Develop theories and practice to build intelligent machines
 Solve real-world problems using AI ideas about representing knowledge,
using knowledge, and assembling systems
 Emphasis is on system building

CPCS335, Introduction to Artificial Intelligence, CS, FCIT, KAU, 1441H/2020G


19
Science-based AI Goal
 Develop concepts and theories to understand biological intelligent
behaviour
• We test theories of human intelligence by writing programs which emulate
human intelligence
 Determine which ideas about representing knowledge, using
knowledge, and assembling systems explain various sorts of intelligence
 Emphasis is on understanding intelligent behaviour

CPCS335, Introduction to Artificial Intelligence, CS, FCIT, KAU, 1441H/2020G


20
AI Approaches
 Four main approaches have been pursued
1. Cognitive Science based approach
2. Laws of Thought based approach
3. Turing Test based approach
4. Rational Agent based approach
 The ultimate objective of these approaches is to build autonomous
intelligent machines

CPCS335, Introduction to Artificial Intelligence, CS, FCIT, KAU, 1441H/2020G


21
Cognitive-Science Based Approach
 Simulate human-like thinking in machines
 Develop theories and practice to build machines with human-like mind
 Emphasis is on human thinking process
• Only intelligent behaviour is not enough, the program must operate in an
intelligent manner (WHAT vs. HOW)
• A machine which passed the Turing test, does not necessarily mean it is intelligent
• The chess programs are quite successful, but they do not tell much how humans play chess

CPCS335, Introduction to Artificial Intelligence, CS, FCIT, KAU, 1441H/2020G


22
Laws of Thought Based Approach
 Develop systems which think rationally
 Use computational models to better understand human’s mind
• How does mind perceive, reason and act
 Focus is on inference mechanisms that are provably correct and
guarantee an optimal solution
• Socrates is a man. All men are mortal. Therefore Socrates is mortal.
 Goal is to formalize the reasoning process as logical rules for inference

CPCS335, Introduction to Artificial Intelligence, CS, FCIT, KAU, 1441H/2020G


23
Turing-Test Based Approach
 Build machines which should perform tasks which at the moment
humans perform better
 Focus is on action or human-like results rather than on intelligent
behaviour
• Achieve results WHICH humans can achieve without focusing on HOW these are
achieved
 A common approach for evaluating systems, e.g. NLG systems
 Goal is to develop systems that are human like

CPCS335, Introduction to Artificial Intelligence, CS, FCIT, KAU, 1441H/2020G


24
Rational-Agent Based Approach
 Act rationally
• Do the right things to maximize the goal achievement
 Emulate intelligent behaviour in machines
 Focus is on systems that act sufficiently if not optimally in all situations
 It is possible to have imperfect reasoning if the goal is achieved
(contrary to laws of thought)
 Goal is to develop systems that are rational and sufficient

CPCS335, Introduction to Artificial Intelligence, CS, FCIT, KAU, 1441H/2020G


25
Intelligence Test: Turing Test
 An interrogator communicates with a person and a computer
 The interrogator can’t see the person or computer
 The computer tries to fool the interrogator
to believe that it is human
 The person also tries to convince the interrogator
that it is human
 If the computer succeeds in fooling the interrogator, then it
passes the intelligence test – intelligent computer

CPCS335, Introduction to Artificial Intelligence, CS, FCIT, KAU, 1441H/2020G


26
Intelligence Test Example
 Alice chat-bot acts like human and won the Turing award
• A: I am Daniel.
• B: Nice to meet you Daniel.
• A: Do you know Clinton?
• B: He was the second Impeached President of the United States.
• A: Do you like him?
• B: Sure he has never done anything to me.
 Is Alice intelligent?
• Absolutely not, because there is no reasoning component
• ca. 50000 fairly trivial input-response rules
• some pattern matching
• some knowledge
• some randomness!
CPCS335, Introduction to Artificial Intelligence, CS, FCIT, KAU, 1441H/2020G
27
Self Review Questions

1 Artificial intelligence is

A Science and engineering of making intelligent machines 

B Multidisciplinary field 

C Yet to be defined precisely 

D All of the above 

CPCS335, Introduction to Artificial Intelligence, CS, FCIT, KAU, 1441H/2020G Designed by Dr. Abullah Basuhail, KAU, 1441H/2020G
28
Self Review Questions

2 Intelligent behaviour involves

A Ability to act in complex environments 

B Ability to learn from experience 

C Ability to think and reason 

D All of the above 

CPCS335, Introduction to Artificial Intelligence, CS, FCIT, KAU, 1441H/2020G Designed by Dr. Abullah Basuhail, KAU, 1441H/2020G
29
Self Review Questions

3 Turing-test based approach characterizes

A Thinking like human being 

B Acting like human being 

C Thinking rationally 

D Acting rationally 

CPCS335, Introduction to Artificial Intelligence, CS, FCIT, KAU, 1441H/2020G Designed by Dr. Abullah Basuhail, KAU, 1441H/2020G
30
Self Review Questions

4 Laws of thought based approach characterizes

A Thinking like human being 

B Acting like human being 

C Thinking rationally 

D Acting rationally 

CPCS335, Introduction to Artificial Intelligence, CS, FCIT, KAU, 1441H/2020G Designed by Dr. Abullah Basuhail, KAU, 1441H/2020G
31
Self Review Questions

5 Cognitive-science based approach characterizes

A Thinking like human being 

B Acting like human being 

C Thinking rationally 

D Acting rationally 

CPCS335, Introduction to Artificial Intelligence, CS, FCIT, KAU, 1441H/2020G Designed by Dr. Abullah Basuhail, KAU, 1441H/2020G
32
Self Review Questions

6 Rational-agent based approach characterizes

A Thinking like human being 

B Acting like human being 

C Thinking rationally 

D Acting rationally 

CPCS335, Introduction to Artificial Intelligence, CS, FCIT, KAU, 1441H/2020G Designed by Dr. Abullah Basuhail, KAU, 1441H/2020G
33
Self Review Questions
 Hassan has developed a machine translation system which can translate
Arabic documents to English. Hassan claims that his system is intelligent.
How could you validate his claim?

CPCS335, Introduction to Artificial Intelligence, CS, FCIT, KAU, 1441H/2020G


34

You might also like