COMP3057

Introduction to Artificial
Intelligence and Machine
Learning
Lecture 1:
Introduction to Artificial Intelligence and Machine Learning
Lecturer: Dr. UMAIR, Zuneera
Course Material

 Where to download course materials?

 Answer: In HKBU Moodle System:
https://buelearning.hkbu.edu.hk
Course title:
COMP3057 Introduction to Artificial Intelligence
and Machine Learning (Section 1/10001) [2022
S1]
In Today’s Lecture …

 Artificial Intelligence
 Introduction, Turing test
 History
 Application areas
 Examples
 Machine Learning
 Introduction and Types
4
 An example of deepfake technology: in a scene from Man of Steel,
actress Amy Adams in the original (left) is modified to have the face
of actor Nicolas Cage (right)
5
Alpha Go
defeated one
of the best
human players
in Go

6
What is Artificial Intelligence?
 Artificial Intelligence or AI is the
branch of Science which deals with
helping machines find solutions to
complex problems in a more human-
like fashion

 Involve borrowing characteristics

from human intelligence and apply
them as algorithms in a computer
friendly way
What is Artificial Intelligence?
 “[The automation of] activities that we associate
with human thinking, activities such as decision-
making, problem solving, learning…”
(Bellman, 1978)
 “The study of mental faculties through the use of
computational models” (Charniak and McDermott,
1985)
 “The study of how to make computers do things at
which, at the moment, people are better”
(Rich and Knight, 1991)
 “The branch of computer science that is concerned
with the automation of intelligent behavior”
(Luger and Stubblefield, 1993)
What is Artificial Intelligence?
Intelligence: “The capacity to learn and solve
problems”
Artificial Intelligence: It is the simulation of
human intelligence by machines.
 The ability to solve problems
 The ability to act rationally
 The ability to act like humans
The Turing Test
 Turing test, proposed by Turing (1950):
“Computing machinery and intelligence”
 Operational test for intelligent behavior

The "standard interpretation" of the Turing

test, in which player C, the interrogator, is
given the task of trying to determine which
player – A or B – is a computer and which
is a human. The interrogator is limited to
using the responses to written questions to
make the determination
Acting humanly: The Turing Test
Programming a computer to pass the test, the
computer would need to posses the following
capabilities:
 Natural language processing
 Enableit to communicate successfully in English (or
some other languages)
 Knowledge representation
 Store information provided before or during the
interrogation
 Automated reasoning
 Use the stored information to answer questions and to
draw new conclusions
 Machine learning
AI Prehistory
AI is a young field
Inherit many ideas, viewpoints, and techniques from other
disciplines:

 Philosophy
Considering the ideas that mind is in some ways like
a machine
Logic, foundations of learning, language

 Mathematics
Formal representation and proof
Algorithms
AI Prehistory
 Psychology
Strengthen the idea that humans and
other animals can be considered
information processing machines.
 Linguistics
 Computer engineering
1956 Dartmouth meeting:
“Artificial Intelligence” adopted
 Overview: AI Application Areas
 Game Playing
 Automated Reasoning and Theorem
Proving
 Early efforts failed to develop a system,
generating an infinite number of possible
solutions;
 Need the powerful techniques to guide their
solution search
 Expert Systems
 Domain-specific knowledge
 Knowledge form a human expert and coding
it into a form that a computer may apply to
similar problems
Overview: AI Application Areas (Cont’1)
 Natural Language Understanding and
Semantic Modeling
 Most work is done in well-understood, specialized
problem areas
 For example, Winograd’s SHRDLU (1973)
 answer queries such as:
 What color block is on the blue cube?
 Plan actions such as:
 Move the red pyramid onto the green brick.
 Much of current work in natural language
understanding
 Finding
representational formalisms that are general
enough to be used in a wide range of applications yet
themselves well to the specific structure of a given
domain.
Overview: AI Application Areas (Cont’2)
 Planning and Robotics
 Language and Environments for AI
 LISP and PROLOG
 Machine Learning
 Programs to learn on their own, either from
experience, analogy, and examples or by being
“told” what to do
 Learning theory
Current Status of AI
 Healthcare
 Heavy Industry
 Finance
 Aviation
 Education
 Face Recognition
 Self-driving……
Current Status of AI
Sophia - The AI Robot

https://www.youtube.com/watch?v=BhU9hOo5Cuc
•In 2015, Google trained a conversational agent
(AI) that can not only effectively interact with
humans as technical support, but also discuss
moral issues, express opinions and answer general
questions based on facts.
Google Assistant – An AI Example

https://www.youtube.com/watch?v=D5VN56jQMWM
What is Machine Learning
Machine Learning is the subfield of computer
science that gives “computers the ability to learn
without being explicitly programmed’. (Arthur
Samuel, 1959).

“A computer program is said to learn from

experience E with respect to some task T and some
performance measure P, if its performance on T, as
measured by P, improves with experience E.”
(Tom Mitchell, 1997)
Why is Machine learning so
important?
 The iterative aspect of machine
learning is important because as
models are exposed to new
data, they are able to
independently adapt. They learn
from previous computations to
produce reliable, repeatable
decisions and results. It's a
science that's not new – but one
that has gained fresh
momentum.
What is Machine Learning

)
 Estimate the relationship between Y and X’s
(model building)
 Also known as supervised learning
 Dependent variables can be explained by the
values of independent variables
What is Machine Learning
 Machine learning (ML) is the study of computer
algorithms that improve automatically through
experience. Machine learning algorithms build a
model based on sample data, known as "training
data", in order to make predictions or decisions
without being explicitly programmed to do so
What is Machine Learning
Use of data to answer questions

Training Predicting
(data) (output)
Types of Machine Learning
Supervised Learning

 In supervised machine learning, a model makes

predictions or decisions based on past or labeled
data.
 Labeled data: data that are given tags or labels
Supervised Learning Algorithms

 Classification
 Regression
Supervised Learning Algorithms (1)
Classification:
 Classification algorithms helps in dividing the
dataset into classes based on different
parameters.
 An email categorized in classes such as
Spam or Not Spam
Supervised Learning Algorithms (2)

Regression
 It helps in predicting the continuous variables
such as prediction of Market Trends,
prediction of House prices, height vs age in
months in children etc.
Unsupervised Learning
 In unsupervised Learning, the algorithm is
trained using un-labelled dataset.
 A model can identify patterns, anomalies, and
relationships in the input data.
Unsupervised Learning Algorithms
(1)
 Clustering is used to group samples such that objects
within the same cluster are more similar to each other
than to the objects from another cluster.
 Create clusters of similar tweets based on their content.
 Find groups of photos with similar cars.
 Identify different types of news.
Unsupervised Learning Algorithms
(2)
 Dimensionality reduction
 It is a technique that transform the data from high
dimension to low dimension- by removing redundant
information and keeping important information.
Supervised vs Unsupervised
Learning Algorithms
 Machine Learning

Supervised
Learning Unsupervised
Learning

Classification Regression Clustering

● Linear ● K-Means, K-
● Support Vector
Regression Medoids Fuzzy
Machine
● SVR, GPR C-Means
● Discriminant
● Ensemble ● Gaussian
Analysis
Methods Mixture
● Naive Bayes
● Decision Trees ● Neural Networks
● Nearest
● Neural ● Hidden Markov
Neighbor
Networks Model
38
Reinforcement Learning
 Using reinforcement learning, the model can
learn based on the rewards it received for its
previous action.
40
AI, Machine Learning and Deep
Learning

41
In Summary …

https://www.youtube.com/watch?v=J4Qsr93
L1qs
The Dark Side of AI

https://www.youtube.com/watch?v=32i5BwJjB8c
Contact