CSE 3013:

Artificial Intelligence
1

B1 SLOT
 Instructor Details

 Name: Dr. Ilanthenral Kandasamy

 Cabin No: SJT 313 A17 / 310 A19
 Phone: 9498028976
 Email: ilanthenral.k@vit.ac.in
 Open Hours: Tuesday (3:00 pm to 4:00 pm)
 Course Grading – Theory
3

 Weightage is as follows:
 Quiz 1 – 10 Marks (Before CAT I)
 Quiz 2 – 10 marks (Before CAT II)
 DA 1 – 10 marks (Recent Trends / Tool based)
 CAT 1 – 15 marks (in Jan end)
 CAT 2 – 15 marks (in )
 FAT – 40 marks (After April 10th)
 Additional Learning – Extra 10 marks
 Course Grading – (J-Component)
4

 Maximum 5 Members
 Review 1 – Topic selection and Literature Survey –
Before CAT I
 Review 2 – 40% implementation of project – Before
CAT II
 Review 3 – Complete project + Viva – one week
before FAT
 Topics - select from given topics
 Reference Book

 Stuart Russell and Peter

Norvig, Artificial
Intelligence - A
Modern Approach,
Prentice Hall, 3rd
edition, 2011.
 Module 1 – Outline

 Definitions
 Importance of AI,
 Evolution of AI
 Applications of AI
 Classification of AI systems with respect to
environment,
 Knowledge Inferring systems and Planning,
 Uncertainty and towards Learning Systems
 Module 2 – Overview to Problem Solving

 Problem solving by Search,

 Problem space - State space,
 Blind Search - Types,
 Performance measurement
 Module 3 – Heuristic Search

 Heuristic Search
 Types,
 Game playing
 mini-max algorithm,
 Alpha-Beta Pruning
 Module 4 – Knowledge Representation
and Reasoning

 Logical systems –
 Knowledge Based systems,
 Propositional Logic – Constraints,
 Predicate Logic – First Order Logic,
 Inference in First Order Logic,
 Ontological Representations and applications
 Module 5 – Uncertainty and knowledge
Reasoning

 Overview – Definition of uncertainty,

 Bayes Rule – Inference,
 Belief Network,
 Utility Based System,
 Decision Network
 Module 6 – Learning Systems

 Forms of Learning – Types –

 Supervised,
 unsupervised,
 Reinforcement learning,
 Learning Decision Trees
 Module 7 – Expert Systems

 Expert Systems‐
 Stages in the development of an Expert Systems‐
 Probability based Expert Systems
 Expert System Tools
 Difficulties in Developing Expert Systems
 Applications of Expert Systems
13
 Sophia
14

 Sophia is a social humanoid robot

 Developed - Hanson Robotics
 50 facial expressions
 Saudi Citizenship
 Chatbot – Provide illusion - understand conversation
 best categorized as a chatbot with a face
 AI methods including face tracking, emotion recognition,
 robotic movements generated by deep neural networks
 Hold eye contact
 Link 1
 Link 2 - Gender
15
 Moley Robotic Kitchen
16

 Can cook over 100 meals

 Moley kitchen could essentially cook any
downloadable recipe on the internet
 Link 1 - Forbes
17
 Furhat the social robot
18

 Tilts head, smiles, exudes empathy and warmth

 Persuades to interact with it as if it were a person,
 picking up on our cues to strike up a rapport.
 aims to build on our new-found ease talking to voice
assistants like Siri (Apple) and Alexa (Amazon)
 AI in Reel life
19

 Many sci fi movies are based on AI

 I, Robot

 Three Laws
 First Law - A robot may not
injure a human being or,
through inaction, allow a
human being to come to harm.
 Second Law - A robot must
obey the orders given it by
human beings except where
such orders would conflict with
the First Law.
 Third Law - A robot must
protect its own existence as
long as such protection does
not conflict with the First or
Second Laws.
AI (2001)
21
Big hero 6
22
 Baymax, the inflatable
healthcare robot
Chitti the robo 2.0
23
 Interesting Links
24

 https://www.hackerearth.com/blog/artificial-
intelligence/7-artificial-intelligence-based-movie-
characters-now-reality/
 Artificial Intelligence
25

 Homo sapiens-man the wise - our mental capacities

are so important to us.
 Tried to understand how we think; that is, how a
mere handful of stuff can perceive, understand,
predict, arid manipulate a world far larger and more
complicated than itself.
 The field of artificial intelligence, or AI, goes further
still: it attempts not just to understand but
also to build intelligent entities.
 AI is one of the newest sciences.
 What is AI?
26

• No clear consensus on the definition of AI

• John McCarthy coined the phrase AI in 1956
 http://www.formal.stanford.edu/jmc/whatisai/whatisai.
html
Q. What is artificial intelligence?
A. It is the science and engineering of making intelligent
machines, especially intelligent computer programs. It is
related to the similar task of using computers to
understand human or other intelligence, but AI does not
have to confine itself to methods that are biologically
observable.
 27

• AI is a collection of hard problems which can be

solved by humans and other living things, but for
which we don’t have good algorithms for solving.
– e. g., understanding spoken natural language, medical
diagnosis, circuit design, learning, self-adaptation,
reasoning, chess playing, proving math theories, etc.
 Russsell & Norvig: a program that
– Acts like human (Turing test)

– Thinks like human (patterns of thinking steps)

– Acts or thinks rationally (logically, correctly)

 AI currently encompasses a huge variety of subfields,
ranging from general purpose areas, such as learning
and perception to such specific tasks as
 playing chess,
 proving mathematical theorems,
 writing poetry, and
 diagnosing diseases.
 AI systematizes and automates intellectual tasks and
is therefore potentially relevant to any sphere of
human intellectual activity.
 Definitions of artificial intelligence according to eight
textbooks are shown in Figure 1.1.
 These definitions vary along two main dimensions.
Roughly, the ones on top are concerned with thought
processes and reasoning, whereas the ones on the bottom
address behavior.
 The definitions on the left measure success in terms of
fidelity to human performance, whereas the ones on the
right measure against an ideal concept of intelligence,
which we will call rationality.
 A system is rational if it does the "right thing," given
what it knows
 Historically, all four approaches to AI have been
followed.
 As one might expect, a tension exists between
approaches centered around humans and
approaches centered around rationality.'
 A human-centered approach must be an empirical
science, involving hypothesis and experimental
confirmation.
 A rationalist approach involves a combination of
mathematics and engineering
32

 Based on Alan Turing

 Trying to decrypt the
Enigma machine,
which the Nazis use to
send coded messages
Alan Turing
33

 widely considered to be the

father of theoretical
computer science and AI.
 Turing Machine
 Turing Test
 https://www.britannica.com
/biography/Alan-Turing
 Imitation Game / Turing test
34

 Uses the "Imitation Game"

 Usual method
Three people play (man, woman, and
interrogator)
Interrogator determines which of the other two
is a woman by asking questions
 Example: How long is your hair?
Questions and responses are typewritten or
repeated by an intermediary
 Acting humanly: The Turing Test approach

 The Turing Test, proposed by Alan Turing (195O), was

designed to provide a satisfactory operational definition
of intelligence.

he suggested a test based on indistinguishability from

undeniably intelligent entities-human beings.
 The computer passes the test if a human interrogator,
after posing some written questions, cannot tell whether
the written responses come from a person or not.
 Computer Capabilities

 The computer would need to possess the following

capabilities:
 Natural language processing to enable it to
communicate successfully in English.
 Knowledge representation to store what it
knows or hears;
 Automated reasoning to use the stored
information to answer questions and to draw new
conclusions;
 Machine learning to adapt to new circumstances
and to detect and extrapolate patterns.
 Total Turing test

 Turing's test deliberately avoided direct physical interaction

between the interrogator and the computer, because physical
simulation of a person is unnecessary for intelligence.
 However, the so-called total Turing Test includes a video
signal so that the interrogator can test the subject's perceptual
abilities, as well as the opportunity for the interrogator to pass
physical objects "through the hatch." To pass the total Turing
Test, the computer will need
 Computer vision to perceive objects, and
 Robotics to manipulate objects and move about.
 These six disciplines compose most of AI, and Turing deserves
credit for designing a test that remains relevant 50 years later.
 Thinking humanly:
The cognitive modeling approach

 If we are going to say that a given program thinks like a

human, we must have some way of determining how
humans think.
 We need to get inside the actual workings of human
minds. There are two ways to do this: through
introspection trying to catch our own thoughts as they go
by-and through psychological experiments.
 The interdisciplinary field of cognitive science brings
together computer models from A1 and experimental
techniques from psychology to try to construct precise
and testable theories of the workings of the human mind.
 Thinking rationally:
The "laws of thought" approach

 The Greek philosopher Aristotle was one of the first

to attempt to codify "right thinking," that is,
irrefutable reasoning processes.
 His syllogisms provided patterns for argument
structures that always yielded correct conclusions
when given correct premises-for example,
 "Socrates is a man; all men are mortal; therefore,
Socrates is mortal."
 These laws of thought were supposed to govern the
operation of the mind; their study initiated the field
called logic.
 By 1965, programs existed that could, in principle,
solve any solvable problem described in logical
notation.
 The so-called logicist tradition within artificial
intelligence hopes to build on such programs to
create intelligent systems.
 There are two main obstacles to this approach.
 First, it is not easy to take informal knowledge and state it in
the formal terms required by logical notation, particularly
when the knowledge is less than 100% certain.
 Second, there is a big difference between being able to solve a
problem "in principle" and doing so in practice
 Acting rationally: The rational agent approach

 An agent is just something that acts

 But computer agents are expected to have other
attributes that distinguish them from mere "programs,"
such as
 operating under autonomous control,
 perceiving their environment,
 persisting over a prolonged time period,
 adapting to change, and
 being capable of taking on another's goals.
 A rational agent is one that acts so as to achieve the
best outcome or, when there is uncertainty, the best
expected outcome.
 The study of AI as rational-agent design has at least two
advantages.
 First, it is more general than the "laws of thought"
approach, because correct inference is just one of several
possible mechanisms for achieving rationality.
 Second, it is more amenable to scientific development
than are approaches based on human behavior or human
thought be- cause the standard of rationality is clearly
defined and completely general.
 Human behavior, on the other hand, is well-adapted for
one specific environment and is the product, in part, of a
complicated and largely unknown evolutionary process
that still is far from producing perfection.