Week 1 a - Introduction.ppsx this is good ppt

it is
ai and artifiicial

1. 1
By
Engr. Dr. Jawwad Ahmad
Introduction
ARTIFICIAL
INTELLIGENCE &
EXPERT SYSTEMS

2. 2
Today’s Goal
Introduction of the Course Instructor
Introduction of the Course
Basic Relation with Searching Methods,
Optimization Techniques & Machine Learning for
Data/Computer Sciences
Engr. Dr. Jawwad Ahmad

3. 3
Instructor Engr. Dr. Jawwad Ahmad
Engr. Dr. Jawwad Ahmad
Electronic Engineer
 Masters in Telecommunication Engineering
 PhD in Telecommunication Engineering
 Head Telecom in Usman Institute of Technology
 HEC Approved PhD Supervisor
 Member of National Curriculum & Revision Committee (NCRC)
 Author of 18 International & National Journal Articles
 Co-PI of Five US Patent
Author of 13 International Conferences
 Author of Four International Book Chapters
 External Examiner of PhD and Masters at NUST, PF-KIET, IU,
HU, SSUET & MUET

4. 4
INTRODUCTION OF THE COURSE
Engr. Dr. Jawwad Ahmad
Data Science or Data Analysis
Searching Methods
Optimization
Machine Learning
Deep Learning
This course introduces modern searching techniques,
optimization methods and machine learning algorithms for
applications in computer science.
Pre-Requisite:
Basic understanding of
computer programming,
linear algebra,
vector calculus,
numerical analysis, and
probability.
Tools

5. 5
INTRODUCTION OF THE COURSE
Engr. Dr. Jawwad Ahmad
Data Science
Data Science builds mathematical models aimed to extract
and represent knowledge from complex data.
It draws techniques from diverse fields, such as, Statistics,
Machine Learning, Data Mining, Information/Signal
Visualization.
It draws expertise from different disciplines, such as,
Statistics, Mathematical Optimization, Computer
Science, Information Technology.

6. Engr. Dr. Jawwad Ahmad 6
Introduction
Data everywhere!
Google: processes 24 peta bytes of data per day.
Facebook: 10 million photos uploaded every hour.
YouTube: 1 hour of video uploaded every second.
Twitter: 400 million tweets per day.
Astronomy: Satellite data is in hundreds of PB.
The Digital Universe of Opportunities
Rich Data and the Increasing Value of the Internet of
Things.
Data comes in different sizes and
also flavors (types):
 Texts
 Numbers
 Graphs
 Tables
 Images
 Transactions
 Videos
Wherever we go, we are “Datafied”
 Smartphones are tracking our locations.
 We leave a data trail in our web browsing.
 Interaction in social networks.
 Privacy is an important issue in Data
Science.
Internet of Things (IoT)
Lots of data
Lots of computation
Various types of communication
 machine-to-machine
 machine-to-human
Smile, we are ‘DATAFIED’!

7. Engr. Dr. Jawwad Ahmad 7
The Data Science Process
Data Mining focuses using machine
learning, pattern recognition and
statistics to discover patterns in data.

8. 8
BIRD EYE VIEW
Engr. Dr. Jawwad Ahmad
Data and Learning Algorithm

9. 9
BIRD EYE VIEW
Engr. Dr. Jawwad Ahmad
Data and Learning Algorithm
y = -1 means
No/False
Learning
Algorithm
Where h is the hypothesis (Decision Box)

10. 10Engr. Dr. Jawwad Ahmad
DATA SCIENCE
ARTIFICIAL
INTELLIGENCE
MACHINE
LEARNING
DEEP
LEARNING
GPT

11. 11
FOUR APPROACHES (IDEAS) FOR AI
Engr. Dr. Jawwad Ahmad
Systems that think
like humans
Systems that think
rationally
Systems that act
like humans
Systems that act
rationally
T
H
O
U
G
H
T
B
E
H
A
V
I
O
U
R
HUMAN RATIONAL
Thinking Humanly
•“The exciting new effort to make computers
think ...
 
machines with minds, in the full and
literal sense” (Haugeland, 1985)
•“The automation of activities that we
associate with human thinking, activities
such as decision-making, problem solving,
learning ...” (Bellman, 1978)
Thinking Rationally
•“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)
Acting Humanly
•“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 which, at the moment, people are
better” (Rich and Knight, 1991)
Acting Rationally
•“A field of study that seeks to explain and
emulate intelligent behavior in terms of
computational processes” (Schalkoff, 1990)
•“The branch of computer science that is
concerned with the automation of intelligent
behavior” (Luger and Stubblefield, 1993)
Here a need of Agent is required that will be discussed later in the Course.

12. 12
THE FOUNDATION OF AI
Engr. Dr. Jawwad Ahmad
Philosophy
(It includes laws governing rationalism, dualism, materialism, empiricism, induction etc.)
Mathematics
(Mathematics formalizes the three main area of AI: computation, logic, and probability)
Economics
(Includes Decision Theory, operational research, and Game Theory etc.)
Psychology
(Provides reasoning models for AI and Strengthen the ideas)
Computer Engineering
(AI has also contributed its own work to computer science, including: time-sharing, the
linked list data type, OOP, etc.)

13. 13
THE FOUNDATION OF AI
Engr. Dr. Jawwad Ahmad
Control theory and Cybernetics
(The artifacts adjust their actions to do better for the linear as well as non-linear
environment over time based on an objective function and feedback from the
environment)
Linguistics
(For understanding natural languages different approaches has been adopted from
the linguistic work such as Formal languages, Syntactic and semantic analysis and
Knowledge representation)

14. Engr. Dr. Jawwad Ahmad 14
What is Artificial Intelligence?
What is Intelligence?
What is Artificial Intelligence?
Intelligence is the computational part of the ability to
achieve goals in the world. Varying kinds and degrees of
intelligence occur in people and animals.
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 intelligence.

15. Engr. Dr. Jawwad Ahmad 15
What is Artificial Intelligence?
Artificial Intelligence : some algorithm to enable
computers to perform actions we dene as requiring
intelligence.
Examples:
Search Based Heuristic Optimization
Evolutionary computation (genetic algorithms)
Logic Programming (fuzzy logic)
Probabilistic Reasoning Under Uncertainty
(Bayesian networks)
Computer Vision
Natural Language Processing
Robotics
Machine Learning (ML)

16. 16
AREAS OF AI AND SOME
DEPENDENCIES
Engr. Dr. Jawwad Ahmad
Search
Vision
Planning
Machine
Learning
Knowledge
Representation
Logic
Expert
Systems
Robotics NLP

17. 17
HISTORY OF AI
Engr. Dr. Jawwad Ahmad
AI has a long history
Ancient Greece
Aristotle
Historical Figures Contributed
Ramon Lull
Al Khowarazmi
Leonardo da Vinci
David Hume
George Boole
Charles Babbage
John von Neuman
 As old as electronic computers themselves (c1940)

18. 18
HISTORY OF AI
Engr. Dr. Jawwad Ahmad
Origins
The Dartmouth Conference: 1956
John McCarthy (Stanford) BS and PhD in Mathematics
Marvin Minsky (MIT) BS and PhD in Mathematics
Herbert Simon (CMU) Electrical Engineer and PhD
Political Science
Allen Newell (CMU) BS in Physics and PhD in
Mathematics
Arthur Samuel (IBM) Electrical Engineering
The Turing Test (1950)
“Machines who Think”
 By Pamela McCorckindale

19. 19
HISTORY OF AI
Engr. Dr. Jawwad Ahmad
FutureToday1700’s
Mathematical
Statistics
1943 – The first ANN
1955 – Official term and
academic recognition
1969 – Backpropagation
1996 – Chess victories –
defeating the world champion1958 – Rosenblatt’s
Perceptron
1985 – Rediscovery of Backprop
2012 – AlexNet wins ImageNet
2013 - Today: Deep Learning is
applied almost everywhere!

20. 20
HISTORY OF AI
Engr. Dr. Jawwad Ahmad
ChatGPT - Solves Anything
Dall-E-2 - Generate Art from
TextSynthesia - Create Talking Avatar
Murf - Your Text to Speech
Do Not Pay - AI Lawyer
Jasper AI - Writes Anything
Chatbot Live - Multipurpose Chatbot
Repurpose IO - Aulpost Social Media
Fireflies - Note Taking
Jenni AI - Writes Essays
Tome App - AI Presentation
Timely - Track Time

21. 21
APPLICATION OF AI
Engr. Dr. Jawwad Ahmad
 Although many of these fields are intermingled, but
applications of AI can be broadly classified among the
following:
Industry/ Robotics
Medical and Health
Online and Telephone customer service
Transportation
Telecommunication
Toys and games
News, publishing, and writing
Natural Language Processing (NLP)
Marketing , Finance, Fraud detection, Money Laundering
etc.

22. 22
Mathematical Modelling
Engr. Dr. Jawwad Ahmad
Black Box
x(t)
h(t)
n(t)
d(t)
y(t)
e(t)
Channel / Plant
d(t) = x(t) * h(t) + n(t)
w(t)
Channel / Plant / System Identification

23. 23
Mathematical Modelling
Engr. Dr. Jawwad Ahmad
Cost / Loss
Function
e(t)
min
e[n]
J[n]=
2
Minimum Mean
Square Error
(MMSE)
E[e
2
(n)]
y = x
2
(Parabola or Convex)

w
o
Optimum
Weights
Channel / Plant / System Identification

24. 24
Surfaces of Cost/Fitness Functions
Engr. Dr. Jawwad Ahmad
Non-Convex , Convex or Concave Surfaces
Need: Lost / Cost Function for
Minima
Need: Fitness Function for
Maxima
Gradient Descent (also often called
 
Steepest Descent)

25. Engr. Dr. Jawwad Ahmad 25
Traditional Vs AI Programming
Computer
Program
Data Output
Computer
Program
Data
Output
Coefficients
Traditional Programming
With Learning Algorithm
Training Phase/Mode
Computer
Program with
Trained
Coefficients
Different
Data
Output

26. 26Engr. Dr. Jawwad Ahmad
Search Algorithms
Uninformed Search
Depth First
Breadth First
Uniform Cost
Informed Search
Greedy
A*
Graph
Games & Adversarial
Search

27. 27Engr. Dr. Jawwad Ahmad
Optimization
Methods
Deterministic
Techniques
Convex
Optimization
Non-Convex
Optimization
Gradient-
Based
Gradient Free
Stochastic
Based
Techniques
Heuristics
(Trajectory
Based)
Metaheuristic
s (Population
Based)
Stochastic
Learning
Techniques
Supervised
Learning
Unsupervised
Learning

28. 28Engr. Dr. Jawwad Ahmad

29. 29Engr. Dr. Jawwad Ahmad
Artificial
Intelligence /
Machine Learning
Classification /
Clustering
Discrete
Output
Data Mining or
Indexing
Recognition
Yes / No
Function Approximation /
Curve Fitting / Regression

30. 30
Summarized Course Outline
Engr. Dr. Jawwad Ahmad
Searching Informed - Uninformed Functions
Gradient Descent, Sub-gradient Descent, Stochastic Gradient
Descent (Like LMS, LMF, NLMS, etc.)
Heuristics (Trajectory Based Like PSO, ACO, etc.)
Metaheuristics (Population Based GA, DE, etc.)
 Artificial Neural Network
 Regression / Curve Fitting / Function Approximation
 Classification / Clustering
Vanishing Gradient (Deep Learning)

31. Engr. Dr. Jawwad Ahmad 31
Thank you