1 Introduction to Business Analytics

Data Analytics, AI, Machine Learning, and so on

1-2

2
Why Data Analytics

¨ Data and analytics are flourishing thanks to…

¤ vast amounts of data
¤ computational power
¤ sophisticated algorithms

¨ Data and analytics capabilities have changed the way

businesses make decisions.
¤ Companies need data-savvy professionals
¤ Turn data into insights and action

1-3

Overview of Business Analytics

¨ Business analytics (data analytics) involves extracting

information and knowledge from data.
¤ Improve the bottom line
¤ Enhance the customer experience
¤ Develop better marketing strategies
¤ Deepen customer engagement
¤ Enhance efficiency and reduce expenses
¤ Identify emerging markets
¤ Mitigate risk and fraud
¨ Business analytics is widely applied.
¤ Marketing
¤ Human resource management
¤ Economics
¤ Finance
¤ Health, sports, and politics
1-4
Overview of Business Analytics

¨ Interdisciplinary
¤ Statistics + Computer Science + Information Systems

¨ Business analytics requires …

¤ Data science knowledge + Business Domain knowledge

¨ Business analytics combines qualitative reasoning with

quantitative tools.
¤ Identify key business problems
¤ Translate data analysis into decisions
¤ Improve business performance
1-5

Overview of Business Analytics

¨ Business analytics begins with understating the business

context.
¤ Ask the right questions
¤ Identify the appropriate analysis
¤ Communicate information

¨ Numerical results are not very useful unless they are

accompanied with clearly stated actionable business
insights.

1-6
Three types of Analytics

¨ There are three different types of analytics techniques.

¤ Descriptive analytics: what has happened?
¤ Predictive analytics: what could happen in the future?
¤ Prescriptive analytics: what should we do?

¨ Turning data-driven recommendations into action also

requires thoughtful consideration and organizational
commitment beyond developing descriptive and predictive
analytical models.

1-7

Overview of Business Analytics

1-8
Descriptive Analytics

¨ Descriptive Analytics: what has happened?

¤ Gather
¤ Organize
¤ Tabulate
¤ Visualize
¤ Summarize
¨ Descriptive information can be presented in a number of
formats.
¤ Written reports
¤ Tables
¤ Graphs
¤ Maps
1-9

Descriptive Analytics

¨ What has happened?

¨ Referred to as business intelligence (BI).
¤ Access and manipulate data through reports, dashboards,
application and visualization tools
¤ Uses past data integrated from multiple sources
¤ Inform decision-making and identify problems and solutions
¨ Examples
¤ A firm’s marketing expenses and sales
¤ Financial reports
¤ Crime rates across regions and time

1-10
Predictive Analytics

¨ What could happen in the future?

¤ Use historical data to make predictions
¤ Build models that help an organization understand what might happen
in the future
¤ Use statistics and data mining
¨ Examples
¤ Identifying customers who are most likely to respond to specific
marketing campaigns
¤ Transactions that are likely to be fraudulent
¤ Incidence of crime at certain regions and times

1-11

Prescriptive Analytics

¨ Prescriptive Analytics: what should we do?

¤ Optimization and simulation algorithms to provide advice
¤ Operations Research/ Management Science
¤ Decision Science
¤ Use statistics and data mining
¨ Examples
¤ Scheduling employees’ works hours
¤ Select a mix of products to manufacture
¤ Choose an investment portfolio

1-12
Data

¨ An important first step for making decisions is to find the

right data and prepare it.
¤ Compilation of facts, figures, or other content
¤ Numerical and non-numerical
¤ All types and formats are generated from multiple sources
¤ Often we have a large amount of data
¤ Even small data can give insights
¨ Data that have been organized, analyzed, and processed in
a meaningful and purposeful way become information.
¨ Use a blend of data, contextual information, experience,
and intuition to derive knowledge.
1-13

Sample Data

¨ It is not feasible to collect data that comprise a population

of all elements of interest.
¤ Too expensive
¤ It is impossible
¨ A sample is a subset of the population and is used for
analyses.
¨ Traditional statistical techniques use sample information to
draw conclusions about the population.

1-14
 Types of Data

¨ Cross-sectional data
¤ Collected by recording a characteristic of many subjects at the same
point in time
¤ Recording a characteristic of many subjects at the same point in time
¨ Time series data
¤ Collected over several time periods focusing on certain groups of
people, specific events, or objects
¤ Hourly, daily, weekly, monthly, quarterly, or annual observations

1-15

Cross-Sectional Data

1-16
Time Series Data

1-17

Structured Data

¨ Structured data
¤ Reside in a pre-defined, row-column format
¤ Spreadsheet or database applications

¤ Enter, store, query, and analyze

¤ Numerical information that is objective and not open to

interpretation
¨ Historically, companies relied mostly on
structured data.
¤ High cost to store and process
¤ Performance limitations
1-18
Unstructured Data

¨ Unstructured data
¤ Do not conform to a pre-defined, row-column format
¤ Textual
¤ Multimedia content
¤ Do not conform to database structures
¨ Human- or machine-generated
¤ Structured human: price, income, retail sales
¤ Structured machine: sensors, speed cameras, web server logs
¤ Unstructured human: email, text, social media, presentations
¤ Unstructured machine: satellite images, video data, camera images

1-19

Big Data

¨ Businesses generate and gather more and more data at an

increasing pace: Big Data.
¤ A massive volume of structured and unstructured data
¤ Extremely difficult to manage, process, and analyze using traditional data
processing tools
¤ Present great opportunities to gain knowledge and game-changing
intelligence
¨ Does not imply complete (population) data
¨ Big data may not be used when available
¤ Inconvenient and computationally burdensome
¤ Benefits may not justify costs

1-20
 Vs of Big Data

¨ There are three characteristics of big data. (3V)

¤ Volume: immense amount of data compiled for a single or
multiple sources
¤ Velocity: generated at a rapid speed, management is a
critical issue
¤ Variety: all types, forms, granularity, structure or
unstructured
¨ Additional characteristics (5V)
Veracity: credibility and quality of the data, reliability
¤
¤ Values: methodological plan for formulating questions,
curating the right data, and unlocking hidden potential
¨ More data doesn’t guarantee the better outcome. 1-21

Categorical and Numerical Variables

¨ A variable is a characteristic of interest that differs in kind or degree

among various observations (records).
¨ Two types of variables
¨ Categorical (Qualitative)
¤ Represent categories
¤ Labels or names to identify distinguishing characteristics
¤ Coded into numbers for data processing
¤ Arithmetic operations are not meaningful (such as zip code)
¤ Example: marital status
¨ Numerical (Quantitative)
¤ Represent meaningful numbers
¤ Arithmetic operations are meaningful
¤ Discrete: assumes a countable number of values (ex, number of children in a family)
¤ Continuous: assumes an uncountable number of values within an interval 1-22
(ex, investment returns)
 Nominal and Ordinal Variables

¨ Four major scales:

1. Nominal
¤ Categorical
¤ Values differ by label or name
¤ Example: marital status
2. Ordinal
¤ Categorical
¤ Reflect labels or name, but can be ranked
¤ Cannot interpret the difference between the ranked values
¤ Example: reviews from 1 star (poor) to 5 stars (outstanding)

1-23

Interval and Ratio

3. Interval
¤ Numerical
¤ Categorize and rank, differences are meaningful
¤ Zero value is arbitrary and does not reflect absence of characteristic
¤ Ratios are not meaningful
¤ Example: temperature
4. Ratio (numerical)
¤ Numerical
¤ Most sophisticated
¤ A true zero point, reflects absence of characteristic
¤ Ratios are meaningful
¤ Example: profits
1-24
Types of Variable - Example

¨ Example: The owner of a ski resort gathers data on tweens.

¨ Music: nominal
¨ Food quality: ordinal
¨ Closing time: interval
¨ Own money spent: ratio
1-25

Data Sources

¨ 90% of the data in the world today was created in the last
two years.
¨ Data sources for this book mostly come from Google.
¤ Bureau of Economic Analysis
¤ Bureau of Labor Statistics
¤ Federal Research Economic Data
¤ U.S. Census Bureau
¤ National Climatic Data Center
¤ Yahoo Finance
¤ Zillow

1-26
Data File Formats

¨ There are standard file formats.

¤ Fixed-width format: each column starts and ends in the same place in every
row
¤ Delimited format: a delimiter separates fields, typically a comma (csv file)
¨ There are three widely used markup languages.
¤ Extensible Markup Language (XML): structured data, each piece enclosed in
a pair of tags, gives information on what the data are
¤ HyperText Markup Language (HTML): structured data with tags, gives
information on how to display the data
¤ JavaScript Object Notation (JSON): alternative to XML, transmit human-
readable data in compact files, not as verbose as XML, supports wide range
of data types, parsing is faster

1-27