Week 1

Taufique Hossain
 LARGE-SCALE DATA IS EVERYWHERE!

§There has been enormous data growth in

both commercial and scientific databases Cyber Security
E-Commerce

due to advances in data generation and

collection technologies
§New mantra
§Gather whatever data you can whenever and
wherever possible. Traffic Patterns Social Networking: Twitter

§Expectations
§Gathered data will have value either for the
purpose collected or for a purpose not
envisioned.
Sensor Networks Computational Simulations
 WHY DATA MINING AND ANALYTICS? COMMERCIAL VIEWPOINT
Lots of data is being collected
and warehoused
­ Web data
­ Yahoo has Peta Bytes of web data
­ Facebook has billions of active users
­ purchases at department/
grocery stores, e-commerce
­ Amazon handles millions of visits/day
­ Bank/Credit Card transactions

Computers have become cheaper and more powerful

Competitive Pressure is Strong
­ Provide better, customized services for an edge (e.g. in Customer Relationship
Management)
BUSINESS ANALYTICS: DEFINITION
Business analytics is the process of examining and analyzing raw data sets to:

● Draw conclusions

● Derive more information

● Improve businesses,
products, and services

Business analytics, is the use of data, information technology, statistical analysis, quantitative
methods, and computer-based models to help managers make better, fact-based decisions.
 BUSINESS ANALYTICS:AREAS

Financial Performance
Analytics Analytics
Risk Analytics
Supply Chain
Analytics

Industry-focused
Customer Analytics
Analytics

Business Analytics Domains

 BUSINESS ANALYTICS:BENEFITS

Cost Reduction Improved efficiency

Good resource
Effective decision
utilization
making

Good market
insights
 TYPES OF ANALYTICS
The three main types of analytics based on the workflow and requirements of data analytics:

What will happen?

Descriptive Prescriptive
Analytics Analytics
Predictive
Analytics

What happened? How can we make it happen?

 DESCRIPTIVE ANALYTICS

● Descriptive analytics is ● This is known as the

designed to accessinformation “simplest class of

about the past. analytics.”

● It focuses on thesummarized ● Break a big chunk of data into

view of facts. smaller pieces, extracting relevant
information from the data of what
happened.
 PREDICTIVE ANALYTICS
Predictive analytics is used in:

• Predicting future outcomes in terms of

probability of an event to occur
• Forecasting weather, plan-failure prediction,
and travel products recommender system
• A predictive model is built on the preliminary
descriptive analytics stage.
 Predictive Analytics Tools

Machine learning algorithms such

as random forests, SVM and
statistics.

Trained data scientists and

machine learning experts building
these models

Popular tools for predictive

analytics: Python, R and
RapidMiner.
 PRESCRIPTIVE ANALYTICS

Prescriptive analytics: the use of optimization

techniques to advise businesses on what they should do

• Prescriptive analytics provides the solution for a

prediction in the future.

• It has the power to suggest favorable solutions and

ease the decision-making process.

Image Source: medicine:Ted.com

Optimization: the process of refining a company’s

efforts to maximize their business outcomes
RETAIL MARKDOWN DECISIONS

Most department stores clear seasonal inventory by reducing

prices.

The question is:

When to reduce the price and by how much?

Let’s do a matching pair to determine which analytics task falls

under which type of business analytics
SCOPE OF BUSINESS ANALYTICS

Descriptive analytics: examine historical data for similar products (prices,

units sold, advertising, ...)

Predictive analytics: predict sales based on price

­ 3 sales scenarios at different price levels

Prescriptive analytics: find the best sets of pricing and advertising to

maximize sales revenue
 Data Analytics LifeCycle

Operationalize Discovery
Deliver final reports, briefs, Learn about business
codes, and technical domain and assess
documents. available resources.

6 1

Communicate Results Data Preparation

Identify key findings, business 5 Data Analytics 2
Execute ELT (extract, load, and
values, and develop narratives transform).
for stakeholders.

4 3

Model Building Model Planning

Develop data sets for testing, Identify techniques and data

training, and production. to understand variables
relationship.
Data Analytics Benefits: Decision-Making

● Companies use business analytics toenable

faster and facts-based decisionmaking.

● Data-driven organizations make better

strategic decisions.

● Companies enjoy high operational efficiency,

improved customer satisfaction, robust profit
and revenue level.
Data Analytics Benefits: Decision-Making

Data analytics helps you define your target

audience based on:

● Customer age group

● Customer preferences

● Location-based purchases

● Popular brands or products people seek

 Data Analytics Benefits: Decision-Making

Data analytics in e-commercehelps:

● Manage inventory

● Forecast demands

● Identify shopping seasons

● Analyze customer sentiment

● Decide optimum prices

Data Analytics Benefits: Cost Reduction

● Data analytics helps understand shopper

behavior by monitoring their browsing
interest.
● Seller identifies shopping pattern and
customer demand.

● Customer data helps companiesminimize

failed campaigns and reduce cost
associated with them.
 Data Analytics Benefits: Cost Reduction

Data analytics helps inreducing

marketing and logistics cost.

Marketers use technologiesto Marketing campaigns use

evaluate customer behavior measured activities to
andmake strategic decisions. plan campaigns.

Predictive analytics is used forbetter

performance, higher ROI, andfaster
success.
 Use of Predictive Analytics for Logistics Management

Predictive analytics helps companies in logistics management by:

● Analyzing current and historical facts to make predictions

● Procuring products based on purchase history

● Organizing customers from shopping patterns and demographic details

● Planning inventory and offloading excessstock

 Factors to Consider in Logistics Planning

Seasons Economic conditions

Weather

Data provides several insights such as identifying products that people tend to buy in a particular season.
 Case Study: Amazon

Amazon uses data analytics to improve efficiency and reduce cost.

Predictive analytics helps to:

Predict what youbuy Anticipate shipping

Such predictions help increase sales and reduce shipping, inventory, and
supply chain costs.
 Case Study: Amazon

● Amazon has morethan 200 fulfillment centers

worldwide.

● Supply chain and logistics optimizationhelp

companies reduce cost and improve
performance.

● Amazon uses data analytics for choosing the

warehouse closest to the customer and reduces
shipping costs by 10–40percent.
Case Study: Amazon

● Amazon uses data analytics to attract customers and

increase profits by an average of 25 percent annually.

● Prices are set based on customer activity on a

website, competitors’ pricing, andproduct availability.

● Product prices typically change every 10 minutes as

data is updated and analyzed.

● Amazon typically offers discounts onthe best-selling

items and earns larger profits on less popular items.

Structured Data
It is the data that is processed,
stored, and retrieved in a fixed
format.
Example: Employee details,
job positions, and salaries.
Types of Data
Unstructured Data
It is the type of data that lacks
any specific form or structure.
Example: Email
Semi-Structured Data
It is the data type containing
both structured and
unstructured data.
Example: CSV and JSON
documents
 Analyzing Unstructured Data

Unstructured information is
About 80% of business data is text-heavy and contains data
unstructured. such as dates, numbers, and
facts.

Internally generated information is Unstructured data is primarily

considered unstructured as the used for BI and analytics but not
intelligence doesn’t fit neatly into a for transaction processing
database. applications.
 Analyzing Unstructured Data

Retailers and manufacturers analyze unstructured data to:

● Improve customer relationship management processes

● Enable targeted marketing

● Perform sentiment analysis on product reviews

The line between unstructured and semi-structured data is not clearly defined.
Unstructured data has some level of structure in it.
 Qualitative and Quantitative Data

Qualitative Data
Data in which classification of objects is
based on attributes and properties.
Example: Softness of skin etc.

Quantitative Data
Data can be measured and expressed
numerically.
Example: Your height and shoe size.
 Qualitative and Quantitative Data

Qualitative Data Quantitative Data

● Data collection is unstructured. ● Data collection is structured.

● It asks why. ● It is all about how much or how many.

● It cannot be computed as it is non- ● It is statistical and is about numbers.

statistical.
● It recommends the final course of
● It develops initial understanding and action.
defines the problem.
 Subgroups of Qualitative Data

Qualitative
Nominal data Ordinal data
Data
Unordered data to which an order is Ordered data that is assigned to
assigned in relation to other named categories in a ranked fashion
categories
Example: Grade classification like pass or fail Example: Feedback to a product with 1–5
for student's test results. ranking.
 Subgroups of Quantitative Data

Discrete data Quantitative Continuous data

Data

It can only take certain values. It can take any value within a
specified range.
Example: The number of students in a
class Example: Share price of a company
 Data Levels of Measurement

It is a classification that describes the nature of information within the values assigned to variables.

Ratio

Interval

Ordinal

Nominal
 Data Levels of Measurement

Nominal Ordinal Interval Ratio

• In nominal level of measurement, numbers in the variable are used to

classify data.
• At this level, words, letters, and alphanumeric symbols can be used.
• Data placed in categories according to a specified characteristic

M F
• Categories bear no quantitative relationship to one another

• Example: People in female gender category are classified as F and those

in male gender are category classified as M.
• customer’s location (America, Europe, Asia)
• employee classification (manager, supervisor, associate)
 Data Levels of Measurement

Nominal Ordinal Interval Ratio

• Ordinal level of measurement depicts ordered relationship among the variable’s

observations.
• Data that is ranked or ordered according to some relationship with one another
• No fixed units of measurement
• It indicates an order of the measurements.
• Interval may not be same
• Examples:
• Runner rankings
• Survey responses – satisfaction (poor, average, good, very good, excellent)
• A student with 100% score is assigned the first rank, another student with
95% score would be assigned the second rank, and so on.
 Data Levels of Measurement

Nominal Ordinal Interval Ratio

• The interval level of measurement classifies and orders the

Temperature in centigrade
measurements.
• Ordinal data but with constant differences between
observations
• No true zero point, can be negative
• It also specifies that the distances between each interval on the
scale are equivalent.

• Examples:
- Example: Temperature in centigrade where the distance
between 80 degrees and 100 degrees is same as the
distance between 1000 degrees and 1020 degrees. 80°C - 100°C = 1000°C - 1020°C
 Data Levels of Measurement

Nominal Ordinal Interval Ratio

● In the ratio level of measurement, observations can have a value of zero.

● Although properties of ratio measurement are similar to the interval level of measurement, the zero in scale makes it
different from the other levels of measurement.
● Ratios are meaningful
● Never fall below zero
Examples:
- monthly sales
- delivery times
 Data for Business Analytics

Classifying Data Elements in a Purchasing Database

 Data for Business Analytics

Classifying Data Elements in a Purchasing Database

A: Categorical E: Ratio I: Interval

B: Categorical F: Ratio J: Interval
C: Categorical G: Ratio
D: Categorical H: Ratio