You are on page 1of 20

Introduction to

Artificial intelligence
Presented by: Mona Nashaat, Ph.D.
2

Agenda

What is AI?

Why AI?

How to understand AI?

Machine Learning?

Types of ML

The ugly truth

Cool stuff we can make with AI


Artificial intelligence
Mimicking Human Intelligence

Artificial intelligence AI
produced by human
the ability to acquire AI is an area of computer
beings rather than
and apply knowledge science that emphasizes the
occurring naturally
and skills creation of intelligent
machines that work and react
like humans
AI is everywhere
Mimicking Human Intelligence

Social Media Sending emails Smart assistants

Travel apps Online shopping


Secure Banking
Why AI?
• The AI market forecast predicts $119 billion
of annual revenue by 2025.
• Wearable artificial intelligence, such as the
Apple Watch, is forecast to expand the
global wearable artificial intelligence
market size to $180 billion by 2025.
• The number of businesses using AI
services grew by 270% from 2015 to 2019.

Netflix’s recommender outranks Siri and Cortana


= $1B revenue annually. with its navigation accuracy
Artificial intelligence
Mimicking Human Intelligence
Computer Science

AI
Symbolic Learning

Machine Learning
Robotics Computer vision

NN (Deep Learning)

Pattern recognition

Statistical Learning

Speech recognition NLP


7

• Machine learning is about extracting knowledge


from data.
• It is a research field at the intersection of
Machine Learning statistics, artificial intelligence, and computer
science.
• It is also known as predictive analytics or
statistical learning.
8

Types of Machine Learning

Supervised learning Unsupervised learning Reinforcement Learning


9

Supervised Learning
• An algorithm learns from example data (training data) and associated
response (target). It is done to predict the correct response when given a new
example (test data).
10

Supervised Learning

Classification Regression
Will the customer renew his Predict weather reading
subscription?

What kind of flower is that?

Rose Lily Tulip


11

Supervised Learning
12

Supervised Learning: Applications


Algorithms: Linear Regression, Logistic Regression, Support Vector Machine,
Decision Tree

Weather Prediction Sales Forecasting


Stock Price Analysis
13

Unsupervised Learning
In unsupervised learning, only the input data is known, and no known output
data is given to the algorithm.
14

Unsupervised Learning: Applications


Algorithms: K Means Clustering, Principal Component Analysis

Customer Segmentation Anomaly Detection


15

Reinforcement Learning
a machine learning training method based on rewarding desired behaviors
and/or punishing undesired ones.
16

Reinforcement Learning: Applications


a machine learning training method based on rewarding desired behaviors
and/or punishing undesired ones.

Robots Gaming Industries


17

The Ugly Truth of Machine Learning

Cost of Data More failures than successes


85% of Machine Learning Projects Fail

Cost of Maintenance Need human experience


18

It is not Magic, it is AI

Hand Gesture Recognition Football Player Detection


19

Try it yourself: ML as a service

Google Cloud Auto AI Huawei ML Kit


20

Thanks
monanashaat@eng.psu.edu.eg

You might also like