If you are in enterprise sales or someone who wants to learn about the

fundamentals of Data Science, this course is right for you.

INTRODUCTION TO DATA SCIENCE

 1. What Data Science is

2. Data Science in Enterprises

3. Why Data Science Matters

4. Use Cases

TOPICS

2
 LEARNING OBJECTIVES
• Describe how Data Science has
evolved in the last 50 years.
• Define four data analysis
techniques.
• State why Data Science is so
important today.
• Identify three to five practical
uses of Data Science.
The purpose of this course is to enable
you to participate in conversations around
Data Science more confidently. Your
learning objectives highlight the main
topics in the course and guide your
learning.

3
 “ Most of the world will make
decisions by either guessing or
using their gut. They will be
either lucky or wrong.”
Suhail Doshi, CEO, Mixpanel

Let’s take a moment to read a quote by Suhail Doshi. Suhail’s words point to a
situation that many businesses find themselves in. The world is awash with data
and those who take advantage of it will be able to make sound business
decisions. On this note, let’s begin our Data Science learning journey.

4
 Our IoT lifestyles and business models have created an explosion of data. With data storage being
relatively inexpensive, much of that data is saved somewhere. And with advanced algorithms and fast
compute at our fingertips, we are now able to mine massive amounts of data quickly and accurately. This
means we don’t need to make decisions on the basis of guesswork and luck. Let’s begin by exploring the
evolution of Data Science and how it became a respected method of extracting knowledge and insights
from data. We’ll hear from someone who uses it day in and day out.

What Data Science is

5
 So…
Data Science is so many things combined. It’s hard to provide a succinct
definition of it

Everyone is talking about it.

Many are claiming to do it.
More and more people are hiring for it.
But what exactly is Data Science?

6
 DATA SCIENCE IS...
A multidisciplinary field that uses scientific methods, processes,
algorithms and systems to analyze data, gain insights and derive value.
More descriptively, it is:
• an opportunity to use technology to process and alter any
volume of digital data, especially big data.

• how businesses gain deeper insights and information

(anything of value) into data assets.

• used to solve specific business problems.

• showing such an enormous range of possibilities that nobody

can afford to ignore it.

• evolving so fast that it is constantly being redefined.

And remember that Data Science isn’t just for large

enterprises. Even small companies can profit greatly from
insights gained by analyzing data. 7
THIS IS ONE WAY TO THINK ABOUT DATA SCIENCE
It’s about data AND MORE
Broadly speaking, Data Science is
in the intersection of all these
domains and more.

Machine Some common data tasks:

Programming Learning Math and
Skills Statistics • Data Mining
• Knowledge Extraction
Data • Information Discovery
Science • Information Harvesting
Research • Data Archeology
• Data Pattern Processing
This venn diagram describes the
Subject different domains that come together to
Matter create the field of Data Science. Data
Expertise Science is at the intersection of
Computer Science, Math & deep
subject matter expertise.
8
 DATA SCIENCE IS MULTIDISCIPLINARY
‘Data Mining’, which grew in importance towards the end of the 20th century, refers to the
process of discovering useful information in large repositories of data. This practice gained
traction rapidly and gave way to the broader term ‘Data Science’ as we know it now.

Currently, Data Science is relevant to any of

the domains shown in the diagram on the
right and is inspired by:

• Problem Solving
• Business Analysis
• Business Strategy
• Domain Knowledge
• Communications
• Presentation
• Inquisitiveness
Here is another attempt to capture the diversity of domains that
contribute to the field of Data Science. And many of these domains
are themselves rapidly evolving, contributing to ever more
sophisticated tools to Data Science. 9
 DATA SCIENCE TIMELINE
It’s not a new profession, but how did it become the hottest job of the 21st century?

Key Milestones
1960s 1970s 1980s 1990s 2000 - today

1962 - mathematician International Computing started Growing interest in 2000 onwards -

John W. Tukey Association for evolving at exponential finding useful patterns Structured efforts to
predicted effect of Statistical Computing rate in data to inform and leverage data for
modern-day electronic announced their inspire those in a business advantage
computing on data mission to link 1981 - IBM released the position to make a
analysis as empirical statistics, computer first PC difference Sale of Internet-connect
science technology and domain devices and mobile
for converting data 1983 - Apple released technology accelerated
1964 - first desktop into knowledge the first PC with the need for a
computer, Programma graphical user generation of experts
101, presented to interface who could manage large
public at New York’s digital data collections
World Fair offered Companies now have
rudimentary analytical the ability to collect
approaches data more easily than
ever before 10
 Companies use different types of analytics based on where they are in their
Data Science journey and the problems they are solving for. Here, you’ll learn
to categorize these methods of data analysis.

Data Science in Enterprises

11
 WHAT WE CAN DO WITH DATA
. . . and why we want to do it
Different techniques that data scientists use for analyzing data.

DESCRIPTIVE & DIAGNOSTIC PREDICTIVE PRESCRIPTIVE

ANALYTICS ANALYTICS ANALYTICS

• Interpreting historical data to • Using mathematics and • Similar to Predictive

understand changes that computational techniques to Analytics but goes one step
have happened in a business. uncover patterns in the data further to find the best
in order to make predictions course of action for a given
• Answers the questions “What about unknown future situation.
happened?” and “Why did it events.
happen?” • Answers the question “How
• Answers the question “What can we make it happen?”
• Retrospective in nature will happen?”
• Prospective in nature
• Prospective in nature

The difference between Predictive & Prescriptive is subtle, and the term predictive analytics is
often times used to represent both. Machine Learning is the core technique used for both.
12
 GARTNER’S CATEGORIES
Higher difficulty of analytics = greater value derived from the analysis
Businesses are motivated to apply
data analytics because the insights
How can we make it
they gain from the analyses help them happen?

make better business decisions and What will

Prescriptive
Analytics
solve problems more effectively. happen?
Predictive
Why did it
Analytics
The data science journey begins with happen?

Diagnostic Foresight
the availability of information and

VALUE
What Analytics
ends with the optimization of happened?

processes, strategies, and solutions. Descriptive

Insight
Analytics

Gartner:
“The more challenging the business Hindsight
problem, the more complex the
analytics will be.”
DIFFICULTY 13
 DATA ANALYTICS
Common Examples

DESCRITIVE/DIAGNOSTIC
PREDICTIVE ANALYTICS PRESCRIPTIVE ANALYTICS
ANALYTICS

• Dashboards • How customers • Suggest the right

• Score Cards respond to a discount to offer on
• Alerts marketing promo an eCommerce
• How sales may be website
Tells what happened in
affected by certain • Tell a truck driver
the past. Doesn’t
market conditions where to stop for
provide much insight
gas
into what will happen Uses past data to model
in the future, or how to future outcomes. Uses techniques like
influence it. optimization or
A/B Testing to advice
managers and workers
on how best to execute
their tasks.

Have a look at some practical applications of data analytics, and think about how each example on this page,
corresponds to these three categories. 14
Why Data Science Matters. industries that leverage Data Science
to solve business problems. why businesses are starting to care
more and more about Data Science.

Why Data Science Matters

15
Data Science is used across
industries for a variety of use
cases. As new applications
are being discovered,
virtually no industry can
afford to ignore the business
advantage that Data Science
offers.

16
WE CARE ABOUT DATA
SCIENCE BECAUSE IT . . .
• Forecasts demands & optimizes supply chains
• Improves user satisfaction
• Generates new discoveries
• Automates complex tasks
• Recommends content
Improves
• Text efficiency
goes here
Gains
• Text customers
goes here
Identifies
• Text fraud
goes here
Predicts
• Text failures
goes here
• Saves lives Data Science helps companies save money through efficiency gains, open
up new markets, automate common tasks, prevent adverse events and
With so many concrete benefits in store,
more. More importantly, it helps companies make better informed
why would any business pass on
decisions, which is sometimes referred to as business intelligence. In
Data Science? short, companies that don’t take advantage of Data Science are not likely
to survive. 17
 Practical applications of Data Science and Data Analytics
how a sales dashboard can help a manager understand where the team’s performance
stands compared to their sales quota for the quarter. The manager can see how this
quarter compares with past quarters. Notice how the analyses draw on historical and
current data to answer questions such as “What has been happening?” and “Why might
these things be happening?” The software used here is Tableau

Use Cases
18
 SALES DASHBOARD Purpose: How are we doing this quarter?
Launch an interactive dashboard https://public.tableau.com/app/profile/mirandali/viz/Salesforce-SalesPerformance/SalesPerformance

NVIDIA CONFIDENTIAL. DO NOT DISTRIBUTE. 19

 DEMAND FORECASTING
Purpose: How can we optimize our supply chain to avoid wastage?

Here is a supply chain

example using NVIDIA
GPUs. This Walmart use
case shows that by
accurately predicting
customer demand,
businesses can optimize
their complex supply chains,
leading to significant cost
savings.

NVIDIA CONFIDENTIAL. DO NOT DISTRIBUTE. 20

In any business, knowing what
content to recommend to a RECOMMENDER SYSTEMS
customer is a challenging task. Purpose: How can we keep our viewers engaged?
How helpful our recommendations
are and how effective they are in
generating business depends on
how well we understand our
customers’ needs and preferences.
Netflix and Amazon, for example,
rely heavily on data and
recommender systems to keep
their customers interested and
coming back for more. In fact,
analytics is so important for
Netflix’s business that they
famously organized a competition
with a $1 million prize, to any data
science team that can improve the
algorithms that they use to
generate these recommendations.
NVIDIA CONFIDENTIAL. DO NOT DISTRIBUTE. 21
 EARLY WARNING SYSTEMS
Early warning Purpose: How can we improve Patient Care?
systems like this one
deployed at Ochsner
is an example of
predictive analytics
used to improve
patient care. Here an
AI tool scans
thousands of data
points in real time to
identify patients in the
hospital that are at
risk for a sudden
cardiac arrest in the
next few hours. This
allows doctors and
nurses to intervene
early and improve
patient outcomes.

NVIDIA CONFIDENTIAL. DO NOT DISTRIBUTE. 22

 AUTONOMOUS VEHICLES
Purpose: Transform the transportation system

If you want a glimpse of how limitless the future of Data Science is likely to be, just take a look at how autonomous vehicles are
pushing the boundaries of data analytics. Autonomous vehicles represent perhaps the most advanced implementations of prescriptive
analytics, where the system doesn’t just prescribe the next step, but actually takes action on its own, in a continuous feedback loop.
We still have some ways to go, but with more advanced algorithms and faster compute, we are getting closer to this dream everyday.
NVIDIA CONFIDENTIAL. DO NOT DISTRIBUTE. 23
 You should now be able to:

1. Describe how Data Science has evolved in the last 50 years

2. Define four data analysis techniques

3. State why Data Science is so important today

4. Identify three to five practical uses of Data Science

We started by addressing the broad question “What is Data Science?”, then went on to talk about its role in enterprises. We
then went on to discuss why Data Science is so important to us today and finished off by looking at several use cases.

Bringing it Together
24
 LINKS TO EXTERNAL RESOURCES

1. Click on the link below to see some more definitions of data science:
https://www.kdnuggets.com/2016/10/battle-data-science-venn-diagrams.html

2. To learn more about the definitions of data analytics, click the following link:
https://halobi.com/blog/descriptive-predictive-and-prescriptive-analytics-explained

3. For more about Walmart’s story, click the link below.

https://www.datanami.com/2019/03/22/how-walmart-uses-gpus-for-better-demand-forecasting/

4. Click the link below to read Ochsner’s story.

https://news.microsoft.com/transform/ochsner-ai-prevents-cardiac-arrests-predicts-codes/

25