Lecture 01

Introduction

Eng. Eslam Mohamed

Agenda
➔ Course introduction and policy
➔ Artificial intelligence , Data science, Machine learning and Deep learning
➔ Process of building an AI system
➔ Different Job fields
➔ Application and use cases
➔ Review for different types of Machine Learning
➔ Classification Vs Regression
➔ Linear Regression
➔ Gradient Descent and Cost Function
➔ Implementation of Linear Regression
The Prerequisites of Artificial intelligence
➔ Strong hold on mathematics ⇒ Linear Algebra, Calculus and Statistics.

➔ Experience with programming languages as Python and/or C++.

➔ A good knowledge in understanding and creating algorithms.

➔ Technical reading and writing which is essential in any technology. (self

learning)
2. Artificial intelligence, Data science,
Machine learning & Deep learning
 Artificial Intelligence is a technique which
enables machines to mimic human behavior
What is Artificial Intelligence?
➔ Artificial intelligence (AI) refers to a set of algorithms that makes machine
acts more intelligently without the need of being explicitly programmed;
This means that it tries its best to mimic the human behavior in learning
and making decisions. The ideal characteristic of artificial intelligence is
its ability to rationalize and take actions that have the best chance of
achieving a specific goal. (Making computers Think and Act like Humans).

➔ The 3 New Powers of Today’s Technology:

◆ Explosion of Data
◆ Exponential Increase in Computing Power and GPU’s
◆ Algorithms Advancement
One Working Definition of AI

Artificial intelligence is the study of how to make

computers do things that people are better at or would
be better at if:
• they could extend what they do to a World Wide
Web-sized amount of data and
• not make mistakes.
AI Purposes
"AI can have two purposes.
One is to use the power of computers to augment
human thinking, just as we use motors to augment
human or horse power. Robotics and expert systems
are major branches of that.
The other is to use a computer's artificial
intelligence to understand how humans think. In a
humanoid way. If you test your programs not merely
by what they can accomplish, but how they accomplish
it, they you're really doing cognitive science; you're
using AI to understand the human mind."
- Herb Simon
A model of knowledge-based systems development

Real
World Problem Reasoning
System
Analysis
Problem
Solution

Representation
Subfields of Artificial Intelligence
➔ Machine learning

➔ Deep Learning (Neural Networks)

➔ Computer Vision

➔ Natural Language Processing

➔ Speech Processing
The Advantages of Artificial intelligence
➔ Reduction in human error.
➔ Takes less time than human due to its multiprocessing capabilities.
➔ Zero risks for humans as AI robots can do the tasks instead.
➔ Availability 24x7 with the same performance unlike human-being who
needs breaks and their performance degrades with time and effort.

➔ Helps in repetitive tasks which could be boring for human (donkey-work).

➔ Daily applications as Google’s assistant or Siri which is irreplaceable
by humans.
The Disadvantages of Artificial intelligence
➔ High cost for both software which needs continuous updating to meet
latest requirements and hardware which needs maintenance and
replacing in case of failure (machine life span)
➔ Increase of unemployment as humans are replaced by bots in many tasks
which leads to the death of many job opportunities such as chat customer
service which is replaced by chat bots.

➔ Machines have no emotions which as it is good for giving unbiased

decisions but bad in making connection as humans make to have a
teamwork.
ECP 512 – Practical Machine Deep Learning
The Obstacles facing Artificial intelligence
➔ The data availability as machine so far learns from the data to take
intelligent decisions based on its training on it. However data is presented
in isolation inside companies or across them.

➔ Data can have problems in inconsistencies and being of low quality.

➔ Lack of qualified professionals and skills that can effectively deploy and
operate AI solutions. This is mainly because of the short timeline of this
field.
The Obstacles facing Artificial intelligence
➔ Cost and time of generating AI solution (fully implemented) as such
solutions take time to be implemented and deployed with a high cost of
the engineers’ salaries, updating software and hardware maintenance.

➔ Models misbehaving due to biasing in the training process.

➔ Interaction issues between humans and AI systems which needs raising

awareness of such technologies for the public to be able to deal with.
What is Machine Learning?
➔ 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 class of tasks
T and performance measure P if its performance
at tasks in T, as measured by P, improves with
experience E.”(Tom Mitchell, 1997)
Machine learning

Data
Computer O utput
Machinelearning program
Traditional Programming chart
◾ It’s an application of AI

◾ Computers observe and analyze Data

Computer program
◾ Predict based on previous patterns O utput

Machine learning chart

Machine Learning Vs Deep Learning
➔ Deep learning is a subset of machine learning that mimics the workings of the
human brain in processing data for use in detecting objects, recognizing speech,
translating languages, and making decisions.
➔ Traditional ML requires manual feature extraction/engineering, but Deep
learning can automatically learn features in data.
➔ Deep Learning is suitable when it comes to problems that need large data as It can
take a huge amount of data--millions of images, for example-- and recognize
certain characteristics. Text-based searches, fraud detection, spam detection,
handwriting recognition, image search, speech recognition, Street View detection,
and translation are all tasks that can be performed through deep learning.
Machine Learning Vs Deep Learning
Quiz
➔ Suppose your email program watches which emails you do or do not
mark as spam, and based on that learns how to better filter spam.
What is the task T in this setting?
● Classifying emails as spam or not spam.
● Watching you label emails as spam or not spam.
● The number (or fraction) of emails correctly classified as
spam/not spam.
● None of the above—this is not a machine learning problem.
Quiz(Solution)
➔ Suppose your email program watches which emails you do or do not
mark as spam, and based on that learns how to better filter spam.
What is the task T in this setting?
● Classifying emails as spam or not spam.
● Watching you label emails as spam or not spam.
● The number (or fraction) of emails correctly classified as
spam/not spam.
● None of the above—this is not a machine learning problem.
Quiz
➔ Which of these is a reasonable definition of machine learning ?

● Machine learning is the science of programming computers

● Machine learning is the field of study that gives computers the

ability to learn without being explicitly programmed

● Machine learning doesn’t learn labelled data

Quiz(Solution)
➔ Which of these is a reasonable definition of machine learning ?

● Machine learning is the science of programming computers

● Machine learning is the field of study that gives computers the

ability to learn without being explicitly programmed

● Machine learning doesn’t learn labelled data

3. Process of building an AI system
 Build An Intelligent AI System

Data Data Data Data Model

Collection Understanding Preprocessing Modeling Deployment
 Build An Intelligent AI System

Data Data Data Data Model

Collection Understanding Preprocessing Modeling Deployment

➔ Gathering and measuring information from

countless different sources
➔ Collecting data allows you to capture a record
of past events so that we can use data analysis
to find recurring patterns
 Build An Intelligent AI System

Data Data Data Data Model

Collection Understanding Preprocessing Modeling Deployment

Ask business –challenge-objective questions to

recognize patterns ,draw conclusions, visualizations
and turns data into insights
 Build An Intelligent AI System

Data Data Data Data Model

Collection Understanding Preprocessing Modeling Deployment

Preparing (cleaning and organizing) the

raw data and transforms it into an
understandable and readable format
 Build An Intelligent AI System

Data Data Data Data Model

Collection Understanding Preprocessing Modeling Deployment

Use Machine Learning Techniques to build

intelligent models to make useful predictions
 Build An Intelligent AI System

Data Data Data Data Model

Collection Understanding Preprocessing Modeling Deployment

Put the intelligent model in production to

be actively use by consumers.
4. Different Job fields
Different Jobs
5. Application and use cases
Applications of AI
➔ Recommender systems – suggesting similar people on
Facebook/LinkedIn. similar movies/ books etc. on Amazon.
➔ Business applications – Customer segmentation, Customer retention,
Targeted Marketing etc.

➔ Medical applications – Disease diagnosis.

➔ Banking – Credit card issue, fraud detection etc.
➔ Language translation, text to speech or vice versa
Applications of AI
➔ Virtual Personal Assistant
◆ Google Now, Alexa are some of the famous examples of virtual
personal assistants. When we ask them something over voice, they
help in finding information.
◆ You should simply initiate them and ask "What is my calendar for
now?", or similar inquiries. We can even instruct assistants for
specific tasks like "Set an alarm for 7 AM next morning” and so on.
◆ The machine learning models behind voice assistants were trained
on human languages and variations in the human voice, because it
has to translate what it hears into words and then make an
intelligent, on-topic response
Applications of AI
➔ People You May Know
◆ Machine learning works on a basic concept that is understanding
with experiences. Facebook continuously notices the profiles that
you visit regularly, your interests, friends that you connect with,
workplace, or a group that you share with somebody etc.
◆ Based on continuous learning, a list of Facebook users are suggested
that you can move forward to become friends with.
Applications of AI
➔ Product Recommendations
◆ Amazon and other online retailers often list “recommended
products” for each consumer individually. These recommendations
are based on past purchases, browsing history, and any other
behavioural information they have about consumers. Often the
recommendations are helpful in finding related items that you need
to complement your purchase (think batteries for a new electronic
gadget). However, most consumers probably don’t realize that their
recommended products are a machine learning model’s analysis of
their behavioural data. This is a great way for online retailers to
provide extra value or upsells to their customers using machine
l.earning
Applications of AI
➔ Email filtering
◆ This is a classic use of machine learning. Email inboxes also have a spam
inbox, where your email provider automatically filters unwanted spam
emails.

But how do they know when an email is spam?

◆ They have trained a model to identify spam emails based on characteristics
they have in common. This includes the content of the email itself, the subject,
and the sender. If you’ve ever looked at your spam inbox, you know that it
wouldn’t be very hard to pick out spam emails because they look very
different from real emails.
Applications of AI
➔ Online Customer Support
◆ Various websites nowadays offer the facility to chat with customer
support while they are exploring inside the site.
◆ But there is no live executive on every website to answer your
queries. We talk to a Chabot in most of the cases. These bots gather
information from the website and present it to the customer queries.
All this is possible due to its machine learning algorithms.
 Applications of AI
➔ Personalized marketing
◆ Marketing is becoming more personal a
◆ Now that much of marketing is online, marketers can use
characteristic and behavioral data to segment the market.
◆ Digital ad platforms allow marketers to choose characteristics of the audience they
want to market to, but many of these platforms take it a step further and continuously
optimize the audience based on who clicks and/or converts on the ads.
◆ The marketer may have listed 4 attributes they want their audience to have, but the
platform may find 5 other attributes that make users more likely to respond to the ads.
6. Review for different types of Machine Learning
Much formal Definition:
➔ Machine learning is a field of data science that focuses on designing
algorithms that can learn from and make predictions on data.

➔ Machine Learning Types:

- Supervised Learning
- Unsupervised Learning
- Reinforcement Learning
 Supervised learning unsupervised learning

Supervised learning, an AI system is Unsupervised learning, AI system is

presented with data which is labeled, presented with unlabeled, uncategorized data
which means that each data tagged with and the system’s algorithms act on the data
the correct label. without prior training.

50
 Reinforcement learning
Reinforcement learning is a type of machine learning in which a computer learns to
perform a task through repeated trial-and-error interactions with a dynamic environment.

51
Any Questions?