Data Science and Machine

Learning
Sania Azhar
CONTENTS OF THIS
Session
Here’s what you’ll find in this Session:

1. Introduction to Data Science

2. Various DS Discipline

3. Career In DS

4. Introduction to ML

5. Why ML is Future?

6. What a Difference in DS,ML,DL and AI

A Picture Is Worth a
Thousand Words
Definition 01
Data science” is as a process,
 What is DS
● Perhaps one of the best ways to understand
“data science” is as a process, namely that
of using data to understand things. Among
the “things” we would like to better we
can list:
● Your customers
● Your business processes
● How your product is consumed / perceived
/ liked by your customers
 What is DS

The key words here are “data” and “understanding”. Data is the main input, understanding is
the ultimate goal. Thus, data science goes far beyond business intelligence, that often stops at
measuring and presenting what has happened without delving into why it happened or how to
control or affect what will happen in the future. Another simple way to put this is that one lets
the “data tell a story”. After, all humans have always tried to make sense of the world around
them by creating narratives that connect the mere facts in logically coherent — if sometimes
fantastic — ways, e.g. myths and legends.
This brings us to a further wording of the same concept, that I especially like: data science
means transforming data into knowledge, that can guide decisions.
Continue
 Another
Definition 02
Data science as an umbrella term.
umbrella term that groups up several crafts

Artificial
Intelligence Machine Learning ETLs, data cleanup

Efficient data
Data Visualization Algorithms and storage and
Infrastructure retrieval
architectures
What is AI 02
 What is AI

Rather than jumping right away into a definition of AI in an abstract, academic and dry way, we would like to first
review three popular applications of AI. 
1. Self Driving Cars
2. Personalized Content Recommendation
3. Computer vision
1.Self Driving
Cars 01
 What is AI

1. Computer vision
2. Search and planning
3. Decision making under
uncertainty
1.Personalized
Content 02
Recommendation
 What is AI

1. Search
2. Predictive machine learning
 03
1.Computer Vision
 What is AI

1. Object detection
2. Object classification
3. Face recognition
4. Image segmentation
5. Flow detection
 03
1.What is AI?
 What is AI

● Definition: AI is the branch of computer science tasked with the design and construction
of intelligent agents.
● An intelligent agent is in turn any agent that is (mostly):
● autonomous (that is, able to perform tasks in a complex environment without constant guidance
by a user) and
● adaptable (able to improve performance by learning from experience)
● able to sense and efficiently abstract features of its environment (perception)
● able to act “rationally”. In this context rational behavior is any that is directed towards
maximizing a quantitative and objective utility function or performance measure.
 What AI isn’t (yet…)

● Robots that talk with a metallic voice (and that are able to scheme against humans). E.g. Terminator /
Skynet.
● Conscious / feeling machines: No matter how good some machines might become in the near future at
imitating feelings, we shall not forget that we are still far from replicating consciousness in digital
computers. For starters we are not really sure what consciousness entails.
● Able to replace any human at any job. Although there is an increasing number of jobs that are being
effectively automated, there is still a large array of more creative or strategically oriented jobs that won’t be
taken over by machines in the medium to long term or possibly ever.
● General intelligence: able to proactively abstract knowledge and skills from one domain and transfer it. See
the previous section.
● Able to (significantly) update themselves proactively/freely. Despite adaptability being one of the
defining characteristics of AI agents, significant updates that allow them to tackle problems that are
qualitatively different from the ones for which they were originally designed, are still out of their reach.
 03
1.What is ML?
 What is Machine Learning

Defining machine learning is not simple. Machine learning has

many applications and overlaps with several other fields.
Combined with the rapid growth of machine learning as a field,
the boundaries of machine learning can be blurry. We like to
define machine learning as a set of tools for making inferences
and predictions from data. Let's compare inference and prediction
tasks to better understand what machine learning can do.
 What can it do?

● Prediction is about the outcome of future events. For example, will it rain
tomorrow? Inference is more vague because it's about drawing insights. We can
infer the causes of events and behaviors, for example, why it rains. We may get a
combination of factors like the month, humidity, and temperature. We can also infer
patterns, for instance, what are the different types of weather conditions? Such as a
rain or overcast. Ultimately, these tasks can work together, because inferences help
make predictions, however they require different types of machine learning.
 How does it work

● Machine learning methods are taken primarily from statistics and computer science.
Machine learning is extremely powerful because it gives computers the ability to
learn without being explicitly programmed to do so. Meaning, the computer can
learn without step-by-step instructions. Essentially, machine learning learns patterns
from existing data and applies it to new data. For example, it can process archived
emails to learn what spam looks like on its own. Then, using what it learned, it can
detect spam in new emails. For machine learning to be successful, it needs high-
quality data. We'll be learning more about all this as we move through the course!
1.What is ML 03
Model?
 Machine Learning Models

● We've defined machine learning, but what does it look like in practice? The answer
is machine learning models. A machine learning model is a statistical representation
of a real-world process, like how we recognize cats or hourly changes in traffic. A
process is modeled using Data.
● We can enter new inputs into a model to get an outcome.
● For example, if we make a model based on historical traffic data, we can enter a
future date into the model to predict how heavy traffic will be tomorrow afternoon.
● The output can even be the probability of an outcome, for example, the probability
that a tweet is fake. In this course, we'll unveil the "black box" that is the model.
 Another example

If you've ever used a streaming service like Netflix, you've seen how these platforms
personalize their home pages based on what they think you would enjoy watching next.
To achieve this, machine learning is used to come up with recommendations based on
your viewing history, the viewing history of similar users, and any other data that could
help understand what you enjoy
● YouTube Recommendation System
 AI, data science, and machine learning walk into
a bar...
● In the age of big data, AI, data science, and machine learning have become buzzwords that
are often incorrectly used interchangeably. They are all related in some way but
encompass a different range of methods and goals.
Which of the following statements is FALSE?

1. Machine learning methods can be considered AI methods.

2. AI and data science share the same goal, but AI uses methods from computer science
and data science uses methods from statistics.
3. Machine learning has become the most prominent subset of AI.
4. Machine learning is used in data science to make predictions.
Three types of machine learning

Reinforcement Unsupervised 
learning difference lies in their
training data
used for deciding
sequential actions,
like a robot deciding its
path or its next move in
a chess game.

Supervised 
Common
 Training data

Machine learning "learns" patterns from existing data and applies it to new data? We
call this existing data "training data". When a model is being built and learning from
training data, we call this "training a model". This can takes nanoseconds to weeks
depending on the size of the data.
 Supervised learning training data

training data for a supervised learning model. We'd like to train a model to predict whether a patient has heart
disease. We have existing records from patients who've experienced chest pains and been tested for heart
disease. Our target variable is "heart disease", because this what we want to predict.
 Supervised learning training data

● The values "True" and "False" are labels for the target variable, meaning whether it's
true or false that a patient has heart disease. Labels don't have to come in this form -
they can be numbers or categories.
● These rows are the observations or examples that our model will learn from. You
should get as many of these as possible.
● And, these columns are features. Features are different pieces of information that
might help predict the target. Age, cholesterol, and smoking habits are known
factors of heart disease. The magic of machine learning is that we can analyze many
features at once, even the ones we're unsure about, and find relationships between
different features. We input labels and features as data to train the model.
 Supervised learning training data

● Once training is done, we can give the model new input. In our case, a new patient.
● The features are inputted
● the model outputs its prediction.