You are on page 1of 11

DataScience in

" The
financial
Industries
through creditinstitutions use
scoring, predictdata
the science to manage credit
manage the workforce. stock market, detect risk
" The
human fraud, and
analytics resource
to screen department the in
candidates,
reduce employee attrition, and monitor
organization
the uses people
mood of the
productivity. improve employee employees,
engagement and
Manyretailers such as
and techniques Walmart and Amazon use data science
and globalsupplythroughout their business that
chain management.
tools
includes marketing
Samatrix.io
QSearch

Pxprimed -2
Jenkine'
Cn consea in
Data Science For Business
" The
primary
to help objective of the course is
you view data

opportunities business
from data science o
etasie Keulede
perspective.
" How the data can be used
information and knowledge.to gather
" Youshould be able to
transform data
into actionable insights. Vsdon

" Based on the insights, by


wisdom take decision andusing
actions to
create an impact.

Samatrix.io
QSerch

Jenkin'
on cenggla in
Case Study-
" Hurricane
Consider an example from a New YorkFrances
a"Hurricane Frances wason its way,
direct hit on Times story from 2004:
Florida's Atlantic coast.barreling across the Caribbean,
away. in Ben- tonville,
situation ...offered Ark., Residents made for higher threatening
a great executives at Wal-Mart
for one of theirStores decidedground,
but
weapons opportunityA week
M. Dillman, predictive technology. that the far
newest data-driven
with forecastsWal-Mart' s chief ahead of the
officer, pressedstorm's
several based on what information
had
landfall,
her staff Linda
to come up
weeks earlier. Backed
that is stored in by thehappened
trillions
when Hurricane
of bytes' Charley struck
'start predicting Wal-Mart' s data
what's going
worth of shopper history
to warehouse, she felt that the
she put it. (Hays, 2004)" happen, instead of waiting forcompany could
it to happen.' as
Samatrix.io
Q Serch

Expaimed -2
Jenkine'
n coneea in
Case Study Hurricane Frances
Whyin the case of naturalcalamity, the
People, data-driven prediction might be useful.
who are in the path of a hurricane,
water. might be interested to buy bottled
" But this is an obvious point, why
discover this fact. we need data science tools and techniques to
" Wal-Mart executives might be
interested in predicting how the hurricane will
impact the sales so that the company can make the necessary
" There could be some coincidence whereas the
arrangements.
released movie CD went up during the week butsales of a particular newly
the impact on sales was
nationwide not just the areas impacted by the hurricane.
" Ms. Dillman is referring to more useful information than some general patterns.

Samatrix.io
Han
Case Study -
" Using
were notdataobvious.
Hurricane Frances
science, the data analyst team
" By could discover the
By
sit analyzing the huge volume of patterns that
few producte analyst could Wal-Mart
and rush stocksidentify of
data from prior similar
" In the
actual ahead thehurricane'
surge in unusual local
s I demand for
(Hays, 2004) scenario, the landfill.
same thing happened. The New
the stores wouldreported that: ".. the experts mined the
indeed need certain data
York Times
flashlights. products-and not justand
thefound
usual
that
" We didn't know in the
seven times past that strawberry
in arecent their normal sales rate, ahead ofPopTarts
interview. And the pre-hurricane
increase in sales, like
a hurricane,'
top-selling itemMs.wasDillman
beer.""
said
Ske 2t af ex
Samatrix.io
Hare
QSaarch

Pxsu mexc-2
lenking'
on conesle in
Variety of Data Types
" Variety is one of
the basic principles of big
" It is also one of
the four
data.
" The data characteristics of big data.
scientist should be able to manage a variety of
" The
information from data types.
various
to images to videos sources of data from
should be integrated bank
to tweets transactions
data management. for analysis and

GOSamatrix.io
Hae
Q Saarch

*Pxsimed-2
Jenking
Cn conega in
Variety of Data Types
" Based on the business
of data. problem, youmay come across different
. Youwould facets
require different data
to analyze and extract management tools and
results for each flavor of data. techniques
" Certain situations such as
data monitoring traffic data require real-time
such management
and
as data analysis to analysis techniques whereas other situations
determine unsuspected patterns require
massive historic data collection.
"In certain situations, we need to
sources for our analysis.
integrate data from a variety of
Samatrix.io
Q Search
Main Categories of Data
"The main
categories of data are
Structured
Unstructured
" Natural language
" Machine-generated
" Graph-based
" Audio, video, and images
" Streaming

Samatrix.io
Q Search

Pxumetd-2
)
Why Understanding Data Type is
" As soon as a new project is assigned, it is
Important
into the exploration always tempting to jump
of statistical and machine
results faster before applying data science. learning models to get
" However, without
understanding
and energy in implementing the data, you would
the solutions that are notwaste your time
given data type. suitable for the
" Whenever you are assigned, youshould spend time in
analyzing the
data according to the different categories of data.

GSamatrix.io
Pxsuimed-2
Structured Data
"
" Structured data has defined length and format.
Number,dates, strings (such as name and
the examples
"
of structured data. address, etc.), are some of
Structured data is usually stored in a
using a structured query database and can be queried
" language (SQL).
Traditionally
the
such as customercompanies have been collecting data from
enterprise relationship sources
management (CRM) data, operational
" Structured resource planning (ERP) data,and
data is about 20% of the data that financial
we
data.
overall system. currently have in the
Samatrix.io
Search

1019
OcicP

Structured Data - Pros


"The structured data is
the machine language.highly organized. It can be easily understood by
The data stored in relational
easily and quickly searched and manipulated. database can be
" The business users can use
need not understand variousstructured
data relatively easily. They
data types and relationships among
them. It is easy todevelop self-service tools for the business user.
"The relational database has been around for a long time now. Several
advanced tools have been developed and tested to manage
structured data. It offers Data managers a variety of advanced tools
and techniques.

Samatrix.io
Ske 2 af 91
Q Search
Pxpuimexd-2

You might also like