VISIT

THE WORLD OF MACHINE LEARNING

SO YOU WON‘T MISS THE ADVENTURE

By Bemnet Girma
linkedin.com/in/bemnetdev

From AI to Deep Learning

20+ Real world Applications
How top tech companies are using AI?

IN JUST 10 PAGES
 www.linkedin.com/in/bemnetdev

VISIT DAY 1/2

Introduction to Machine Learning

Objectives
• What is machine learning?
• Differences between AI, machine learning Vs Deep Learning?
• Why we should learn machine Learning?
• What are different types of machine Learning?

What is Artificial Intelligence (AI)?

Artificial intelligence (AI) refers to the simulation of human intelligence in machines that are
programmed to think like humans and mimic their actions.

What is Machine Learning?

Machine learning is a subset of artificial intelligence (AI). It is focused on teaching computers to learn
from data and improve with experience (accuracy) – instead of being explicitly programmed to do so.

“
Machine Learning is the science (and art) of giving computers the ability to learn and make
decisions from data without explicitly programmed.
”
What is Deep Learning?
Deep Learning is a subfield of machine learning concerned with algorithms inspired by the structure
and function of the brain called artificial neural networks.

The “Deep” inside Deep Learning?

Deep learning is a branch of machine learning that uses neural networks with many (deep) layers. A
deep neural network analyzes data with learned representations similarly to the way a person would
look at a problem. Deep learning networks will often improve as you increase the amount of data being
used to train them.

In traditional machine learning, the algorithm is given a set of relevant features to analyze a given data.
However, in deep learning, the algorithm is given raw data and decides for itself what features are
relevant.

1|Page
 Visit the world of Machine Learning

Artificial Intelligence Vs Machine Learning Vs Deep Learning

Artificial intelligence is essentially when machines do tasks that typically require human intelligence.
The field of artificial intelligence includes machine learning, where machines can learn by experience
and acquire skills without human involvement whereas Deep learning is a branch of machine learning
where neural networks – algorithms inspired by the human brain – learn from large amounts of data.

Why we should learn Machine Learning?

 Big data explosion
 Hunger for new business and revenue
 Advancement in Machine Learning Algorithms
 Development of extremely powerful machines with high capacity and faster computing ability.
 Storage capacity

Types of classical machine Learning

Based on the amount and type of supervision they get during training, machine learning can be classified
into five types.

1. Supervised Learning is a type of machine learning where well-labelled training data is used to
train the machines.

Definition of supervise: observe and direct the execution of (a task or activity).

 Regression is a supervised learning technique which helps in finding the correlation between
variables and enables us to predict the continuous output variable based on the one or more
predictor variables.

Definition of regress: re (back) + gress (a Latin word meaning step) then

regress mean step back to the best fit line or curve.

Example: House price prediction using a given set of features (like area in square meter).

2|Page
 www.linkedin.com/in/bemnetdev

 Classification is a supervised learning technique which helps in finding the correlation between
variables and enables us to predict class label output variable based on the one or more predictor
variables.

Definition of classify: to consider (someone or something) as belonging to a particular group.

Example: Spam email filtering using flag words

“
The goal of supervised machine learning problem is to find a single model that will best
predict our wanted outcome. Rather than making one model it is better to ensemble models
into account, and average those models to produce one final super model.
”
 Ensemble method is a machine learning technique that combines several base models (weak
learners) in order to produce one optimal predictive model.

Definition of ensemble: a group of items viewed as a whole rather than individually.

2. Unsupervised Learning is a type of machine learning where unlabelled datasets are used to finds
all kind of unknown patterns in data.

 Clustering is a pattern discovery technique that divides data points into a number of groups
without having any prior knowledge of their group memberships rather based on their
similarities.

Definition of cluster: a group of similar things or people positioned or occurring closely together.

Example: Customer segmentation based on their experience

3|Page
 Visit the world of Machine Learning

 Association Rule Learning is a rule-based machine learning technique which is used to find
interesting relationships and associations hidden in large data-sets.

Definition of association: a connection or cooperative link between people or organizations..

Example: Market Basket Analysis based on association between items

“
Large numbers of input features can cause poor performance for machine learning
algorithms. It is a good idea to try to reduce the dimension of your training data using
dimensionality reduction techniques before you feed it to a machine learning algorithm. It
will run much faster, the data will take up less memory and computational time. Moreover it
is easy to visualize features using 2D plots.
”
 Dimensionality Reduction refers to techniques for reducing the number of input variables in
training data without losing too much information.

Example: Suppose if you want to predict a student’s exam result, the number of pages a student
reads may be strongly correlated with how long he/she read, so the dimensionality algorithm
will merge them into one feature that represents the number of pages per minute.

4|Page
 www.linkedin.com/in/bemnetdev

3. Semi-supervised Learning is a type of machine learning that combines a small amount of labelled
data with a large amount of unlabelled data during training.

Example: Some photo hosting services, such as Google photos, are good examples of this. Once
you upload all your family photos to the service. It automatically recognizes that the same
person A shows up in photos 1, 5, and 11, while another person B shows up in photos 2, 5, and
7. This is the unsupervised part of the algorithm (clustering). Now all the system needs is for
you to tell it who these people are. Just add one label per person and it is able to name everyone
in every photo, which is useful for searching photos.

4. Self-supervised Learning is a type of machine learning that doesn’t require human input to
perform data labelling. The results are obtained by models that analyze data, label, and categorize
information independently without any human input.

“
Self-supervised learning exploits unlabelled data to yield labels. This eliminates the need for
manually labelling data, which is a tedious process.
”
The only difference is that, unlike unsupervised learning, self-supervised learning does not
perform the grouping and clustering of data, as is the case with unsupervised learning.

This learning type allows machines to examine part of a data example to figure out the
remaining part. In simple terms, self-supervised learning learns from unlabelled data to fill in
the blanks for missing pieces. This data can be in the form of images, text, audio, and videos.

Example: If you have a large dataset of unlabelled images, you can randomly mask a small
part of each image and then train a model to recover the original image. During training, the
masked images are used as the inputs to the model, and the original images are used as the
labels.

5|Page
 Visit the world of Machine Learning

5. Reinforcement Learning is a feedback-based Machine learning technique in which an agent learns

to behave in an environment by performing the actions and seeing the results of actions. For each
good action, the agent gets positive feedback or reward, and for each bad action, the agent gets
negative feedback or penalty. A Policy defines what action the agent should choose when it is in a
given situation.

Definition of reinforcement: the action of strengthening or encouraging something.

Example: DeepMinds’s AlphaGo program is a good example of Reinforcement learning: it

made the headlines in May 2017 when it beat the world’s champion Ke Jie at the Chinese
popular board game. It learned its winning policy by analysing millions of games, and then
playing many games against itself. Note that the learning was turned off during the game
against the champion; AlphaGo was just applying the policy it had learned.
So these were the most kick-starting fundamentals of machine learning. Which of these topics or
machine learning types have you been interested?

Key takeaways
• AI refers to simulation human intelligence in machines.
• ML is the study of training computers to learn from data without being explicitly programmed.
• DL is the study of training computers to learn like a human brain.

• Supervised Learning is a type of ML where machines train with well-labelled training data.
• Regression is a supervised learning technique which helps to predict numerical outputs.
• Classification is a supervised learning technique which helps to predict categorical outputs.
• Ensemble method is a ML technique that combines models to produce one optimal model.

• Unsupervised Learning is a type of ML where unlabelled data is used to find unknown pattern.
• Clustering is a pattern discovery technique that divides data points into a number of groups.
• Association is a ML technique used to find relationships & associations hidden in large dataset.
• Dimensionality Reduction is a technique for reducing no. of input variables in the training data.

• Semi-supervised Learning is a type of ML that combine few labelled & many unlabelled data.
• Self-supervised Learning is a type of ML that doesn’t require human input for data labelling.
• Reinforcement Learning is a feedback-based ML technique where agent learn to act in envt.

6|Page
 www.linkedin.com/in/bemnetdev

VISIT DAY 2/2

Machine Learning in a Real World

Objectives
• What are real life applications of machine learning?
• Which popular companies are using machine learning?
• How are we using machine learning in our daily lives?

Real World Applications of Machine Learning

Face Recognition
One of the most common applications of Machine Learning is Automatic Friend Tagging
Suggestions in Facebook or any other social media platform. Facebook uses face detection
and Image recognition to automatically find the face of the person which matches it’s
Database and hence suggests us to tag that person based on DeepFace.

Image Recognition
Can you unlock a phone simply by looking at it? If yes, you are using machine learning. The
high-end camera of your phone recognizes 80 nodal points on a human face and machine
learning technologies to measure the variable of a person’s face and unlock the phone.

Social Media Services

From personalizing your news feed to better ads targeting, social media platforms are
utilizing machine learning for their own and user benefits. Machine learning is the core
element of Computer Vision, which is a technique to extract useful information from images
and videos. Pinterest uses computer vision to identify the objects (or pins) in the images and
recommend similar pins accordingly.

Stock market forecasting

Machine learning is widely used in stock market trading. In the stock market, there is always
a risk of up and downs in shares, so for this machine learning's long short term memory
neural network is used for the prediction of stock market trends.

Price prediction
Uber uses a dynamic pricing model, where the price per trip depends on the surge in demand
and the number of driver availability. The pricing algorithm normalizes the prices once there
is an increase in the availability of the drivers. If you are getting late for a meeting and you
need to book an Uber in a crowded area, get ready to pay twice the normal fare.

7|Page
 Visit the world of Machine Learning

Web search Engine

Google and other search engines use machine learning to improve the search results for you.
Every time you execute a search, the algorithms at the backend keep a watch at how you
respond to the results. If you open the top results and stay on the web page for long, the
search engine assumes that the results it displayed were in accordance to the query. else, it
estimates that the results served did not match requirement. This way, the algorithms
working at the backend improve the search results.

Medical Diagnosis
In the field of healthcare, machine learning & deep learning have shown promising results
in a variety of fields, namely disease diagnosis with medical imaging, surgical robots, and
boosting hospital performance. One such application of deep learning to detect brain tumor
from MRI scan images.
A brain tumor is a mass or growth of abnormal cells in your brain.

Online Fraud Detection

Machine learning is making our online transaction safe and secure by detecting fraud
transaction. Whenever we perform some online transaction, there may be various ways that
a fraudulent transaction can take place such as fake accounts, fake ids, and steal money in
the middle of a transaction. So to detect this, Feed Forward Neural network helps us by
checking whether it is a genuine transaction or a fraud transaction.

Product Recommendation
Suppose you check an item on Amazon, but you do not buy it then and there. But the next
day, you’re watching videos on YouTube and suddenly you see an ad for the same item.
You switch to Facebook, there also you see the same ad. So how does this happen? Well,
this happens because Google tracks your search history, and recommends ads based on your
search history.

Self-driving (Autonomous) car

One of the most exciting applications of machine learning is self-driving cars. Machine
learning plays a significant role in self-driving cars. Tesla, the most popular car
manufacturing company is working on self-driving car. It is using unsupervised learning
method to train the car models to detect people and objects while driving.

Virtual Personal Assistant

As the name suggests, Virtual Personal Assistants assist in finding useful information using
speech recognition, when asked via text or voice. All you need to do is ask a simple question
like “What is my schedule for tomorrow?” or “Show my upcoming Flights“. For answering,
your personal assistant searches for information or recalls your related queries to collect info.

Online Video Streaming

With over 100 million subscribers, there is no doubt that Netflix is the daddy of the online
streaming world. Netflix’s speedy rise has all movie industrialists taken aback – forcing them
to ask, “How on earth could one single website take on Hollywood?”. The answer is Machine
Learning. The Netflix algorithm constantly gathers massive amounts of data about users’
activities use it properly.

8|Page
 www.linkedin.com/in/bemnetdev

Automatic Language Translation

Traveling to a new place is always thrilling but the only enigma is to understand the common
language of that place. To solve this dilemma, Google has launched an app that can help in
the easy translation of any language. Google uses ‘Google Neural Machine Translation’ that
has the ability to absorb thousands of languages, words, and dictionaries and transmute any
sentence in the desired language.

Traffic prediction
If we want to visit a new place, we take help of Google Maps, which shows us the correct
path with the shortest route and predicts the traffic conditions. It predicts the traffic conditions
such as whether traffic is cleared, slow-moving, or heavily congested with the help of two
ways: (1) Real Time location of the vehicle form Google Map app and sensors. (2) Average
time has taken on past days at the same time. Everyone who is using Google Map is helping
this app to make it better. It takes information from the user and sends back to its database to
improve the performance.

Email Spam and Malware Filtering

Whenever we receive a new email, it is filtered automatically as important, normal, and spam.
We always receive an important mail in our inbox with the important symbol and spam emails
in our spam box, and the technology behind this is Machine learning.

Online Customer Support Chatbot

A number of websites nowadays offer the option to chat with customer support representative
while they are navigating within the site. However, not every website has a live executive to
answer your queries. In most of the cases, you talk to a chatbot. These bots tend to extract
information from the website and present it to the customers. Meanwhile, the chatbots
advances with time. They tend to understand the user queries better and serve them with better
answers, which is possible due to its machine learning algorithms.

Video surveillance
With the help of algorithms like behavior analysis, image detection, we can train the
machines to detect crime on the basis of their behavior. The cameras track the behavior of
people and classify them as usual or unusual. And once some unusual behavior is tracked,
the system can then alert the human attendants which could then avoid any mishaps. Thus,
this helps to reduce the crime rate and improves the quality of surveillance.

Building an intelligent bot for a game

DeepMinds’s AlphaGo program is a good example of game bots: it made the headlines in
May 2017 when it beat the world’s champion Ke Jie at the Chinese popular board game. It
learned its winning policy by analysing millions of games, and then playing many games
against itself. Note that the learning was turned off during the game against the champion;
AlphaGo was just applying the policy it had learned. After the match between AlphaGo and
Ke Jie, The self-taught AlphaGo Zero achieved a 100–0 victory against the early competitive
version of AlphaGo, and its successor AlphaZero is currently perceived as the world's top
player in Go.

9|Page
 Visit the world of Machine Learning

Automatically flagging offensive comments on discussion forums

Hate speech is one of the serious issues we see on social media platforms like Twitter and
Facebook daily. Most of the posts containing hate speech can be found in the accounts of
people with political views. Twitter’s apparently faster at automatically flagging and
removing abusive tweets. Twitter is asking users to think twice before sending mean tweets.

Sentiment Analysis
This is an application of natural language processing (NLP) where the words of documents
are converted to sentiments like happy, sad, angry which are captured by emotions
effectively. Amazon is using sentiment analysis for movie and product reviews. Rating (five
stars to one star) are automatically attached as soon as you write your review using sentiment
analysis programs.

Text Classification
Every news website classifies the news article before publishing it so that every time visitors
visit their website can easily click on the type of news that interests them. For example, I like
to read the latest technology updates, so every time I visit a news website, I click on the
technology section the read the headlines. Google news is using natural language processing
to summarizing long news articles to a headline and classify news articles automatically.

Customer Segmentation
Customer segmentation by means of machine learning is a process of dividing a customer
base into particular groups with similar characteristics. There are countless ways to segment
customers. It can be done demographically or psychographically (Age, Sex, Highest level of
education achieved, Stage of life, Income, Religion, Attitudes towards products/services).
Apple is using customer segmentation to provide the best products and services for their
customers.

Demand Forecasting
Market demand forecasting is a critical process for any business, but perhaps
none more so than those in consumer packaged goods. Inventory, production,
storage, shipping, marketing – every retail companies’ operations are affected
by accurate forecasting. Samsung Electronics Austria forecasts demand with
Brightics AI to avoid shortage or overstocking and wasting money in logistics,
maximizing sales opportunities.

So these were some of the most popular examples of machine learning applications in the real
world. Which of these applications have thrilled you?

Key takeaways
• Companies like Google, facebook, amazon… are applying ML to their product and services.
• ML can be applied in the following areas:
Prediction and forecasting
Image (face) and Speech (voice) recognition
Text filtering and classification [NLP]
Recommendation systems
AI Gamming

10 | P a g e