MODULE 1: OVERVIEW OF BIG DATA Business Intelligence in such a way that it

AND BUSINESS ANALYTICS can be analyzed by people. It is comprised

of the tools with which people can
Introduction disseminate the data that’s been collected,
to produce meaningful conclusions
Everyone who runs a business would tell (Fulton, 2013).
you that making decisions is the most
important task in the role. Business 2. Business Analytics is the use of data,
managers are always faced with questions information technology, statistical
on what to sell, how much to sell, to analysis, quantitative methods, and
whom, through what mode. A manager or mathematical or computer-based models
entrepreneur has to know how much it to help managers gain improved insight
would need to make his or her product, about business operations and make
how to cut down on expenses, how to better, fact based decisions (Evans, 2012).
market the product to customers, and
when the business can be expanded. 3. Business analytics is comprised of
Decisions do not just come about on their solutions used to build analysis models
own. They are a product of analysis based and simulations to create scenarios,
on the information available to the understand realities and predict future
decision maker. states including data mining, predictive
analytics, applied analytics and statistics
Business managers have always gathered (https://www.gartner.com/it-glossary/bus
information to base their decisions on. But iness-analytics).
with the development of computers which
can store and process large amounts of 4. Business Analytics is the study of data
data, business analytics began to offer through statistical and operations
more possibilities for decision makers. analysis, the formation of predictive
models, application of optimization
techniques, and the communication of
1.1. Business Analytics and Business these results to customers, business
Intelligence partners, and college executives (Galleto,
2018).
In the era of knowledge economy, getting
the right information to decision makers at 5. Business analytics refers to a broad
the right time is critical to their business use of various quantitative techniques
success, and one such attempt includes such as statistics, data mining,
the growing use of business analytics (Min, optimization tools, and simulation
2017). supported by the query and reporting
mechanism to assist decision makers in
Business analytics is one of the most making more informed decisions within a
talked-about topics in the field of business closed-loop framework seeking continuous
and information technology. And as process improvement through monitoring
expected, business analytics is becoming and learning (Min, 2017).
one of the most sought-after courses in the
academe.
Business Intelligence (BI)
Business Analytics (BA) 1. Business Intelligence as the process of
collecting information from all sources to
1. Business Analytics is the art of make data-driven decisions in an
assembling the data gathered through organization (Fulton, 2013, Galleto, 2018).
2. Business Intelligence as the process of amounts of data being generated
combining aspects of reporting, monitoring every day
and alerting, dashboards, scorecards, and ○ 2005 - Google Analytics

ad hoc query data exploration (Capistrano, ○ 2010 - popularisation of big data &

2018, Galleto, 2018) introduction of big data analytics

1.2. Origin of Business Analytics 1.3. Big Data and Business Analytics

From ancient times, people have had this We have already defined Business
need to predict the future. They have used Analytics and Business Intelligence. This
crude methods like stones and sticks to time, let us learn another concept related
project how much crops they were going to to BA which is the Big Data.
harvest or how much they were going to
yield. In this section, we shall trace the So, what is big data?
development of business analytics from
the ancient to the modern time SAS defines big data as a term that
describes the large volume of structured
1. Prehistoric and unstructured data which can be
● The earliest humans would use sticks and analyzed for insights needed for better
stones to help predict sales trends of new decisions and strategic business moves.
inventions.
● First forms of business analytics may be IBM, on the other hand, refers to it as data
statistics whose uses can be traced back at least sets whose size or type is beyond the
to the biblical times in ancient Egypt, Babylon, ability of traditional relational databases to
and Rome. Attributed to its usefulness for capture, manage, and process the data
helping the policy maker (including ancient
with low-latency.
rulers or kings) make a better decision

2. Industrial (1800s)
● Frederick Winslow Taylor develops the concept
What is the difference?
of business management as a scientific
discipline Business Analytics
● 1908 - Henry Ford would apply this principle - focuses on financial and
and measured the time it took for each operational analytics of the
component of his Ford T Model to assemble. business
- relies on the person looking at the
3. Information age (mid-20th century) data and drawing inferences from
○ Digital Age / 1970 - computers it.
played a huge role in the
advancement of business analytics Big Data
with the introduction of Decision
- helps to analyze a broader range of
Support Systems
Information Age / 1980 - Data
data coming in from all sources
○
warehouses became popular to store
- involves machine automation to
and organize large amounts of data,
analyze data.
the first iteration of today’s server
farms. The importance of big data is not on how
○ 1990 - Introduction of Enterprise much data you have, but what you do with
Resource Planning by Gartner those data.
4. Internet age (21st century)
○ The internet, big data and the cloud
came along, with the massive
There are four aspects that define big data 2. Data warehousing. This is where the
which are volume, variety, velocity and data is cleansed, curated, organized, and
veracity. ready for analysis.
1. Volume is about how huge the data sets IT automatically joins data in a
are. “normalized form” in preparation for the
2. Variety includes how many pieces of business analyst
data from diff sources we gather together
from social media data, government data, 3. Extract, Transform, or Load
financial data, banking data, all sorts of Processes (ETL). This is the process of
transactions all combined together to moving data from source systems to data
make one or more profiles for your warehouse to an analytical tool.
customers.
3. Velocity is the speed of data. How data becomes business value:
4. Veracity means that there is a lot of It has to be extracted, curated and
uncertainty, meaning, there is all of these cleansed, joined in a data warehouse in
different data coming together, but the which point is now prepared for
problem is we don’t know what to do with analytics.
them.
2.2 Types of Analytics
MODULE 2: BUSINESS ANALYTICS Now that we have collected the
FRAMEWORK right data and processed them into usable
form, we need to find at what level are we
Introduction going to dig deep into the data to get the
fact-based insight we need to find out so
we can come up with better business
Information, in its raw form, is not much decisions.
useful for business decision making. It has
to be collected, analyzed, and present in a There are three types of analytics that can
way that is useful for decision making. help businesses make the most out of the
big data that they have and find the
solutions to problems that they have:
2.1 Framework for Business Analytics descriptive, predictive and prescriptive.
The process of turning raw data into 1. Descriptive Analytics. This is the data
business action is the framework for that is used to benchmark or to profile.
Business Analytics. Concerns itself with presenting data as it
is.
There are 3 steps in turning data into
analytics which are Data Extraction, Data Most Common Approaches / Methodologies;
Warehousing and the Extract, Transform,
or Load Processes (ETL). 1. Cross Tabulation
-Is getting granular bits of data. Summarizing and

1. Data extraction. This is the first step in aggregating it for better analysis.

turning data into analytics. There are at -It is the most basic methodology , various analytics build
on cross-tabulation
least 3 sources of data which are the
source systems, raw transactions, and 2. Hypothesis and Inferencing
from documents and forms. Once data is transformed we can draw early insights and
conclusions.
 ● Correlation - type of inference, determine which It is the foundation of predictive analytics. Types of model
factors move together and which factors don’t. and algorithms used depend on what kind one is
● T-test - type of inference , determines the modelling.
similarities or difference between two groups.
Regression - attempt to model independent variables
3. Data Visualization (predictors) to predict a dependent variable (outcome)
After tabulation and inference, best way to show insights ● Linear regression - for continuous variables (a
is through visualizations such as charts: person’s salary).
● Time-Series (Trend) - how value changes ● Logistic regression - for probability or likelihood
overtime of an event (likelihood a person will earn 50k a
● Stacked Area (Composition) - how it changes, month)
intuitive feel of where values are
● Heat Map (Relative Weight) - columns, high and Other predictive analytics model: clustering, decision
low values trees, support vector machines. Neural networks -
● Histogram (Distribution) - to visualize 2 machine learning methodology
population
● Scatterplot (Relationship) - shows relationship To perform predictive analytics, first understand what
between 2 variables. This chart is more relevant kind of outcome you’re modelling and what are the
with predictive analytics. predictors for that outcome.

Tools in Descriptive Analytics: To do a Model…

1. Spreadsheets
Excel, easy tool to generate cross-tabulation via Pivot 1. Training a Model
Table. A model must first be trained. It involves getting the
sample data with the relationship you want to model.
2. Business Intelligence (BI) tools A model is trained by applying a statistical model to a
Combines ease of reporting with point-and-click training data and then surfacing the relationship between
generation of charts for visuals. Allows users to load data predictor and outcome (the shape of the trend line in the
and visualize data immediately. Ways to combine charts, example is THE RELATIONSHIP that is modelled).
MS Power BI, QlikView

3. Custom Data Visualization tools

Used for problems that are too complex for a spreadsheet
and something you can’t immediately do on a
point-and-click. d3js, ggplot, matplotlib

2. Predictive Analytics. This is used to Once the relationship is determined then the model can be

determine relationships between two tested for accuracy. Ex: trend line is tested using
R-squared with a residual graph to see how far or close
different types of data and make the outcome is to the historical data that was trained on.
predictions about future data.
2. Outputting Model Components
Concerns itself with modelling The model components ARE the relationships generated
relationships between predictions and from the statistical model. These components are later
outcomes. These relationships are tested used for predictions. Often more useful to understand the
for accuracy on an initial sample of data. nature of the relationship itself.
Once an acceptable model is generated,
predictions are generated by feeding the
model a new set of prediction data.
Common approaches / methodologies:

1. Statistical Analysis
The coefficients are the model components from the essence of prescriptive analysis: knowing
regression line. the costs and benefits of a decision.

3. Outputting Predictions
MODULE 3: DATA AND DATABASE
With model components determined, predictions can now
be created by inputting new values of the variable into the
MANAGEMENT
model. Predictions are generated by substituting new
values for the test data.
Introduction
Good business decisions rely on access to
accurate data. To facilitate this,
organisations need a system for recording,
updating, and tracking data on a regular
basis. Once collected, data can be
retrieved and presented in formats useable
to managers and employees of the
organization through computerised
databases. Various database elements can
Tools for Predictive Analytics be linked through database management
systems.
1. Spreadsheets
Excel can do simple modelling via regression and solver.
1. Data, Database and Database
Measures in excel can be generated using data analysis
function called Solver
Management Systems

2. Statistical Modelling Workflow Interface

Allows an analyst to easily string together very complex
What is data?
predictive modelling processes. Knime, RapidMiner, SAS, For every database system, the heart of
SPSS Modeler each system is what you call the data
(Recario, 2018).
3. Programming Interfaces
Used if the model is too complex or the relationship is too Data are facts or figures which we can
specific. Allows customization of the modelling to suit store in a database. An example of this is
specific problems. Any computer language is allowed, your ID number, the name of your teacher,
most popular are: SAS, OpenSource, Python etc.. the number of students in your class now.
Summarize - (1) Preparing data - (2) Creating model - (3)
Raw Facts and Figures
Figuring model relationships of data

What is a database?
3. Prescriptive Analytics. This is used to
create recommendations through 1. A database is a collection of logically
simulation and optimization models. It related data and it is typically visualize as
involves providing advice on what actions tables; composed of cells matched with
can be taken given the predictions about several columns and rows (Recario, 2018).
the future and what drives the future. The
2. In computer terms, database is a
collection of data, not necessarily always
stored in a computer, such as records
stored in a filing cabinet, in a notebook
(Crisanto, 2018).
What is a Database Management System update, and delete) our data within
(DBMS)? the DBMS.
● Database definition. Metadata
describes the contents of your
1. A DBMS is a collection of interrelated database. This is a data that
data using the software and hardware describes another data in terms of
used to access the data in a useful manner its constraints, size, and the data
such as database or set of databases type of the tables or the databases
stored in a disk or other media, a that are contained within your
computer, or set of computers where the DBMS.
software to access the database can run ● Data Processing. This allows the
(Crisanto, 2018). user to perform different types of
operations such as string
2. A DBMS is a software package or manipulation, date operations and
software that allows you to store, retrieve, so on. You can also manipulate the
package your database (Recario, 2018). data based on an event or what
they call ‘triggers’ which is based
on the event or the conditions of
2. Functions of a Database Management the event.
System What are the functions of a ● Data sharing. This means that
DBMS? within a database, privileges are
provided to different types of users.
What are the advantages of using a Different types of users can access
DBMS? Here are some of the functions of the same database but the same
a DBMS (Crisanto, 2018, Recario, 2018 ): users can have different privileges.
● The DBMS can provide a
convenient and efficient interface 3. Components
for storing, retrieving, and updating of a Database
data, and for extracting useful System
information from the database, and
a clear and logical view of the
process that manipulates the data. A database
● Data Independence: DBMS system has the
maintains the segregation between following
the program and the data components
● Concurrency Control. This refers (Crisanto, 2018):
to a process to ensure that data is
updated correctly and
appropriately when more than one
person is accessing the data 1. Users
● Recovery Services. This refers to
mechanisms for recovery of data The users of a database system are
that is apparently lost due to the database administrator, system
system crashes developer, and end user.
● Utility Services: An initialization ● A database administrator (DBA) is
and maintenance operations on a the person responsible for all the
database. data resources in an organization.
● Manipulation of data. It tells us The DBA uses tools that come with
that we can CRUD (create, retrieve, a DBMS to improve the
 productivity and performance of The final component of a database
database planning and design. system is the database itself. This is the
● The system developers are those space in the disk or computer where the
groups of people who create the data are actually stored.
application programs that cater to
the user requirements. And, they
use their own tools to write 3.4. Components of a Database
programs that communicate with Management System
the DBMS.
● Finally, the end-users in an
organization are the ones that can
add, update, and delete data in a
database through application
programs or directly through a
DBMS. They use the application
program to accomplish their
day-to-day tasks.
2. Database Application A database system has the following
The next component of the components (Recario, 2018):
database system is the database
application. This is a computer program 1. Interface
that allows users to manipulate the data in
a DBMS through a user-friendly interface. The interface is responsible for
It can be divided into four broad interacting with the user. It is also called
categories: the “middleman” as it communicates
● Personal which is restricted to a between two different parties: the end
single user user, which is you, and the database
● Departmental which is referenced management systems.
by hundreds of users over a shared
system or network The interface allows the end user to input
● Enterprise which extensions of values that will be understood by the
departmental applications involving database management system. It is also
thousands of users responsible for the receiving and
● Internet which is the largest form outputting, and providing the results of
of information sharing where the queries or the commands that the end
billions of users are involved. user issued.
2. Parser
3. Database Management System
(DBMS) The parser is responsible for checking the
The DBMS decouples application correctness of the syntax that was
programs from data. The databases store provided by the end user through the
all its data in one location, thereby limiting interface. So it also checks the tables and
data duplication. other objects in the query.
Examples of DBMS include Access, Oracle, 3. Query Optimizer
IBM’s DB2, and SQL Server.
If the parsing is correct and proper, then it
4. Database will proceed to what we call the query
optimizer.
The query optimizer comes from the root
word ‘optimize’, which means it simplifies Three types of analytics.
the query form or the input into a simpler 1. Descriptive : data-driven
type of input that can be digested and can benchmarking
be understood by the database
management system. This allows the This is data that is used to
DBMS to save time and save space when benchmark and to profile to be able to
performing operations. establish historical trends. It is the
commonly used analytics by companies.
4. Execution Engine Not making any predictions nor looking for
any relationships yet - just looking at data
The next one is the execution engine. Its as it is.
main responsibility is to execute the
output of the query optimizer. Once the 2. Predictive : data driven prediction
query has been optimized, then it will pass
the data or the optimized query to the We now then look at the
execution engine whose sole purpose is to relationships of the data. Once we know
execute the command. the relationships inside the data, we can
make predictions and forecasts by feeding
5. Storage new data into the predictive model.
Storage is the physical location of your
database, which can be the secondary or 3. Prescriptive: data driven
the primary memory. recommendation
The third type of analytics is the
prescriptive analytics. It involves providing
MODULE 4: APPLICATIONS OF advice on what actions can be taken given
BUSINESS ANALYTICS the predictions about the future and what
drives the future. The essence of
Introduction prescriptive analysis: knowing the costs
and benefits of a decision.
We learned that raw data needs to be
extracted from its sources, cleansed,
curated, and warehoused to prepare it for Most common applications of Business
analysis. These data can be analyzed in Analytics in the Industry
different ways depending on the problem
or situation at hand. They can be ● Credit Card Companies - Credit
leveraged in various ways to optimize and debit cards are an everyday
business operations. part of consumer spending.
● Customer Relationship
Three types of commonly used analytics -- Management (CRM) - Excellent
descriptive, predictive and prescriptive. customer relations is critical for
Applications for business analytics any company that wants to retain
especially in the three functions which are customer
mostly present in the industry: Finance, ● Finance - Corporations turn to
Human Resource, and Marketing. business analysts to optimize
budgeting, banking, financial
planning,
● Human Resources - Great
1. Types of Analytics businesses are composed of a great
 staff, and it’s HR’s job to not only Analytics plays an important role in
find the ideal candidates but keep human resources.
them on board.
● Manufacturing - Business analysts ● Suppose you are an HR manager
work with data to help and you want to understand the
stakeholders understand the things demographics of the employees in
that affect operations and the your company. You may need
bottom line. information on the diversity of the
● Marketing - Business analysts employees’ location, address,
help answer these questions and so school graduated, etc. This is
many more, by measuring where descriptive analysis can be
marketing and advertising metrics, used to manage human resources
better.
2. Applications of Business Analytics in ● As an HR manager, you may also
Finance want to determine which employees
are likely to resign, because
Applications of business analytics in recruiting new people will mean
finance industry. additional cost for the company.
You may use predictive analysis
● An example of descriptive analysis in this case by looking at the
in finance is using analytics in historical patterns of resignations
reporting financial results, from to determine the likely causes of
gathering financial inputs from resignations and the number of
different sources, cleansing it, to employees that are likely to resign
reporting it. in the future.
● Predictive analysis, on the other ● Finally, as an HR manager, you
hand, is when we want to predict may also want to determine the
the trend of sales for the next two drivers that make employees stay
months using historical patterns of in the company. Employee
seasonality, and examining engagement, such as looking at
whether investing a lot in sales what makes them content, happy
people might also drive the sales and stay in the company (ex. party,
trend. bonus, free training, etc) is where
prescriptive analysis can be used.
● Prescriptive analysis is applied
when we want to determine the
feasibility of the project, say the 4. Applications of Business Analytics
likelihood that the project will Marketing
falter, or overshoot the budget, or
fail. This is with the assumption Analytics is very useful especially if you
that we have looked into are in the marketing industry.
consideration the various inputs
such as cost of materials, location ● Let’s say you are in charge of
of project, etc monitoring the results of an online
marketing and sales campaign in
your company. You can use
3. Applications of Business Analytics in descriptive analytics to determine
Human Resources (HR) how many people clicked the ads,
how many people bought the
 product, how many people paid
cash-on-delivery, or by credit card.
1. Ethical and Legal Considerations in
● If you want to understand how Business Analytics
factors (ex. price, marketing mix
and attributing the effect, Some examples on ethical and legal
channels, mode of payment, etc) considerations in business analytics such
contribute to the performance to as danger of discrimination and bias,
predict the future performance problem when the integrity of the data has
(success or failure) of a campaign, been compromised, and problems on the
or to achieve targets, you can use lack of transparency. Data privacy law and
predictive analytics. some examples on addressing data privacy
Concerns.
● Prescriptive analytics in
marketing can be seen through the
recommendation engines which are 1.1 Discrimination and Bias
found to be successful in driving
more sales. These are the When we make decisions based on the
recommendations that you can see results of business analysis involving big
whenever you visit an online data, there is a high probability that it will
shopping website, say to buy a affect thousands and even millions of
book. This recommendation will go people.
something like “those who bought
that book also bought this book”. Discrimination and bias happens when
This recommendation is based on our judgement and assumptions are
the prescriptive algorithms which influenced by our own biases. (Bañez,
look at the patterns of people who 2018).
buy the same product or profiles Suppose you used business analytics to
that look similar or items that are screen candidates for employment, and
purchased together. you decided that we are going to eliminate
those who are not employed one year prior
application, since you may feel that this
MODULE 5: ETHICS ISSUES IN seems objective and rational. You want to
BUSINESS ANALYTICS hire people who have continuity in working
and the one-year gap may be a source of
Introduction suspicion. However, this decision may be
biased to certain group of people such as
Ethics is a moral principle that somehow mothers who have to take a one-year leave
guides a person on what is bad and what to take care of their children after giving
is good. We need to be aware of our ethical birth. You may be biased against these
obligations as we practice and apply mothers who will not be qualified for the
business analytics in your profession. The job, not because they are not fit for the
results of business analysis when used as job, but because they chose to take care of
basis to make important decisions can their children after giving birth. These
affect people and the organization, thus, biases will involve serious consequences
there is a need to learn about the ethical and certain people will be unfairly denied
and legal considerations and implications resources.
of business analytics. As business
analytics practitioners, we also need to In this example, your bias against women
know some of the general information may cause a portion of women ending up
about the data privacy being rejected from consideration even
though they are just as good, or even Suppose you think that your company
better, than other candidates. made a wrong decision in not promoting
you. So how do you appeal to this if you
Bias filters out people and limits their resources, think that their method is based on bias,
this limitation to resources then results in risk and thus faulty? To appeal, you need to
factors such as lack of jobs or education. These know the process to determine if you are
people who are devoid of jobs or education are then right, and they are wrong, and decided
filtered out by the bias which creates a based on bias. You also need to know that
self-fulfilling loop. Bias - No Resources - Risk the data sets they used to decide is
Factors - accurate. But chances are, none of these
would be available to you because of
Tools should be validated if it reflects the real
non-disclosure, intellectual property, and
world. Tools are only a means to an end not the end
other institutional agreements you have
with the company. How can you now
itself
question that decision and make your
appeal if you don’t know how the decision
arrived at?
1.2 Integrity of Data Analytics So here, you can see that we need
Another ethical issue concerns the transparency to provide us access as our
integrity of data analytics. As we use the means to understand how systems work
results of analytical analysis to make and affect our lives.
certain decisions, we should also take These analytics can be interpreted as a “black box”
consideration the assumptions used to where information is put, processes it and provides
ensure its validity. Using wrong results without giving information of the internal
assumptions, even if we use the right
workings or what happened in between.
analytical tools will bring us to wrong data.
The right tool is useless if we have
corrupted, incomplete, outdated, or When people are introduced to the inner workings
irrelevant data. of the “black box”, they would most likely encounter
mathematical jargons and get overwhelmed and
An example of this is when a credit card then give up - this term is called “the authority of
company used a data on a client’s inscrutable”
purchases which he never made in the
first place, or mistakenly used on a This system is indeed convenient but opaque.
different client with the same name. It can
also be using correct data but interpreting Principle of transparency requires not just access
it the wrong way. These are just some of but to provide meaningful access - giving means to
the examples that show how integrity understand the essence of how it works.
of data can be compromised.
Practitioners require to be responsible for the 2. Ethical Implications of Business
integrity of their data, methods and interpretations. Analytics
Disclosure of customer data is prevalent in
1.3 Lack of Transparency day-to-day business, especially in the
internet via social media and searches,
Transparency, is another ethical issue and most individuals are unaware that
which is related to the first two issues that whenever you perform a search or post
we have discussed earlier. something in social media, you already
leave a digital trace in the world wide web
and these traces can be used to profile you Not only does the person have the right to
without your consent (Ligot, 2018). You be informed, he/she also has the right to
should know by now that this is true and access the data that a company may have
possible through the use of technology. about him/her--not only the content but
Our information are being collected and the sources of the information as well. You
analyzed, each time we visit a website, can go to the companies that maintain
click an ad, shop online, etc, and most your information and ask them exactly
people are not aware of it. what they keep. The right to information
and the right to access can help address
our earlier concerns about bias and
Data Privacy Law transparency.
With the information you can obtain, you
We leave digital traces whenever we go can determine whether or not the
online, even if we are not disclosing our information was processed fairly or if they
identity. And these traces are used by use correct or updated information. Since
marketing companies to collect and you know the information that they have
process data such as which websites we about you, you can determine whether or
visit, which products we buy, which ads not that information is correct or updated.
we watch, and so on. With the growing
concern that people are losing their ● The Right to Rectify
information from companies who collects
data without their knowledge, has You also have the right to rectify. If the
prompted many countries, including the information they have about you is
Philippines, to pass data privacy laws. inaccurate our outdated, under the data
privacy act, you have the right to have it
In the Philippines, the Data Privacy Act corrected. You have the right to complain if
was made to protect people and bring back the company fails to rectify the data or if
control of their own information. the company fails to respect your other
rights, you may complain before the
As a business analytics practitioner, you National Privacy Commission or you can
need to know some of the general go to courts.
information about the data privacy act
(Bañez, 2018). ● The Right to Complain
Once it is established that your rights are
• Consent Required and the Right to be violated, the company which violated your
Informed rights can be made to pay a fine, and pay
you directly, damages. In some cases, the
Before you can collect and process one’s company’s officers may become criminally
information, you need to get his/her liable – they can go to jail. The company
consent. The consent has to be recorded that processes your data will also have the
and informed, meaning the person has to responsibilities to make sure that your
understand the purpose which his/her data is protected from unauthorized
data is being collected and processed. He access. So, they can’t just leave your data
or she has the right to know how his/her out there in the open – they have to take
data will be processed and whom data will reasonable steps to make sure that your
be shared. data is secure. This responsibility, along
with the right to access and right to rectify
● The Right to Access your information, can help address our
earlier concerns about integrity.
 1.2 Impacts and Challenges
1. Introduction to Analytics Benefits
◦ Reduced costs, better risk management,
Analytics is the use of: faster decisions, better productivity and
● data enhanced bottom-line performance such
● information technology, as profitability and customer satisfaction.
● statistical analysis,
● quantitative methods, and Challenges
● mathematical or computer-based ◦ Lack of understanding of how to use
models analytics, competing business priorities,
to help managers gain improved insight insufficient analytical skills, difficulty in
about their business operations and make getting good data and sharing information,
better, fact-based decisions. and not understanding the benefits versus
perceived costs of analytics studies.
1.1 Applications of Business Analytics
Pricing 1.3 Scope of Business Analytics
◦ setting prices for consumer and
industrial goods, government contracts, Descriptive analytics: the use of data to
and maintenance contracts understand past and current business
performance and make informed decisions
Customer segmentation
◦ identifying and targeting key customer Predictive analytics: predict the future by
groups in retail, insurance, and credit card examining historical data, detecting
industries patterns or relationships in these data,
and then extrapolating these relationships
Merchandising forward in time.
◦ determining brands to buy, quantities,
and allocations Prescriptive analytics: identify the best
alternatives to minimize or maximize some
Location objective
◦ finding the best location for bank
branches and ATMs, or where to service Example:
industrial equipment Most department stores clear seasonal
inventory by reducing prices. Key
Social Media question: When to reduce the price and by
◦ understand trends and customer how much to maximize revenue?
perceptions; assist marketing managers
and product designers Potential applications of analytics:
Descriptive analytics: examine historical
data for similar products (prices, units
sold, advertising, ...)
Predictive analytics: predict sales based
on price
Prescriptive analytics: find the best sets
of pricing and advertising to maximize
sales revenue
 2. Tools Example of Data Sources and Uses
● Database queries and analysis Internal
● Spreadsheets ● Annual reports
● Data visualization ● Accounting audits
● Dashboards to report key ● Financial profitability analysis
performance measures ● Operations management
● Data and Statistical methods performance
● Data Mining basics (predictive ● Human resource measurements
models)
External
● Economic trends
2.1 Software Support ● Marketing research

● SQL - various databases New developments: Web behavior – Social

● Excel - Spreadsheets Media – Mobile - IOT
● Tableau Software - Simple drag ● page views, visitor’s country, time
and drop tools for visualizing data of view, length of time, origin and
from spreadsheets and other destination paths, products they
databases. searched for and viewed, products
● IBM Cognos Express - An purchased, what reviews they read,
integrated business intelligence and many others.
and planning solution designed to
meet the needs of midsize 3.1 Big Data
companies, provides reporting,
analysis, dashboard, scorecard, Big data to refer to massive amounts of
planning, budgeting and business data from a wide variety of
forecasting capabilities. sources, much of which is available in real
● SAS / SPSS / Rapid Miner - time, and much of which is uncertain or
Predictive modeling and data unpredictable. IBM calls these
mining, visualization, forecasting, characteristics volume, variety, velocity,
optimization and model and veracity.
management, statistical analysis,
text analytics, and more using “The effective use of big data has the potential to
visual workflows. transform economies, delivering a new wave of
productivity growth and consumer surplus. Using big
● R / Python - Advanced data will become a key basis of competition for existing
programing-based data companies, and will create new competitors who are
preparation, analytics and able to attract employees that have the critical skills
visualization. for a big data world.” - McKinsey Global Institute,
2011

3. Data 3.2 Datasets and Databases

Data: numerical or textual facts and Database - a collection of related tables
figures that are collected through some containing records on people, places, or
type of measurement process. things.
◦ In a database table the columns
Information: result of analyzing data; that correspond to each individual element of
is, extracting meaning from data to data (called fields, or attributes), and the
support evaluation and decision making. rows represent records of related data
elements.
Data set - a collection of data (often a
single “spreadsheet” or data mining table).
◦ Examples: Marketing survey responses, a
table of historical stock prices, and a
collection of measurements of dimensions
of a manufactured item.
3.3 Types of Data
Discrete - derived from counting Data Reliability and Validity
something.
◦ For example, a delivery is either on time Reliability - data are accurate and
or not; an order is complete or incomplete; consistent.
or an invoice can have one, two, three, or Validity - data measures what it is
any number of errors. Some discrete supposed to measure.
metrics would be the proportion of on-time
deliveries; the number of incomplete Examples:
orders each day, and the number of errors ◦ A tire pressure gage that consistently
per invoice. reads several pounds of pressure below
the true value is not reliable, although it
Continuous - based on a continuous scale is valid because it does measure tire
of measurement. pressure.
◦ Any metrics involving dollars, length, ◦ The number of calls to a customer service
time, volume, or weight, for example, are desk might be counted correctly each day
continuous. (and thus is a reliable measure) but not
valid if it is used to assess customer
dissatisfaction, as many calls may be
Measurement Scales simple queries.
◦ A survey question that asks a customer
Categorical (nominal) data - sorted into to rate the quality of the food in a
categories according to specified restaurant may be neither reliable
characteristics. (because different customers may have
conflicting perceptions) nor valid (if the
Ordinal data - can be ordered or ranked intent is to measure customer satisfaction,
according to some relationship to one as satisfaction generally includes other
another. elements of service besides food).
Interval data - ordinal but have constant
differences between observations and have 4. Models
arbitrary zero points.
Model - an abstraction or representation
Ratio data - continuous and have a of a real system, idea, or object.
natural zero. - Often a simplification of the real
thing.
- Captures the most important
features.
- Can be a written or verbal
description, a visual
representation, a mathematical
formula, or a spreadsheet.
4.1 Three Forms of a model ◦ To add prior knowledge about the
relationship between variables.
The sales of a new product, such as a
first-generation iPad or 3D television, often The task of the modeler is to select or build
follow a common pattern. an appropriate model that best represents
the behavior of the real situation.
1. Verbal description: The rate of sales
starts small as early adopters begin to Example: economic theory tells us that
evaluate a new product and then begins demand for a product is negatively related
to grow at an increasing rate over time as to its price. Thus, as prices increase,
positive customer feedback spreads. demand falls, and vice versa.
Eventually, the market begins to become
saturated and the rate of sales begins
to decrease. Linear Demand Prediction Model
2. Visual model: A sketch of sales as an As price increases, demand falls.
S-shaped curve over time Issues: Demand can become negative +
empirical data has a poor fit.

3. Mathematical model:
S = aebect Non-Linear Demand Prediction Model
where Assumes price elasticity is constant
• S is sales, (constant ratio of % change in demand to
• t is time, % change in price)
• e is the base of natural logarithms, and
• a, b and c are constants that need to be
estimated.

*from data to model slide #27*

4.2 Model Assumptions

Assumptions are made to..
◦ To simplify a model and make it more
tractable; that is, able to be easily analyzed
or solved. 4.3 Uncertainty and Risk
Uncertainty is imperfect knowledge (of Identify the price that maximizes profit,
what will happen in the future). subject to any constraints that might exist.
Risk is the potential of (gaining or) losing max. Profit - s.t. Sales >= 0 - Sales is
something of value. It is the consequence integer
of actions taken under uncertainty.
Often measured using standard deviation 5. Problem Solving with Analytics
of variables. (=Deviation risk measure)
1. Recognize a problem
“To try to eliminate risk in business enterprise is futile.
Risk is inherent in the Problems exist when there is a gap
commitment of present resources to future
expectations. Indeed, economic between what is happening and what we
progress can be defined as the ability to take greater think should be happening.
risks. The attempt to
eliminate risks, even the attempt to minimize them, can For example, costs are too high compared
only make them
irrational and unbearable. It can only result in the with competitors.
greatest risk of all: rigidity.”
2. Define the problem
– Peter Drucker
Clearly defining the problem is not a trivial
task. What is part of the problem? What
4.4 Prescriptive Decision Models not?
Prescriptive decision models help Complexity increases when the following
decision makers identify the best solution. occur:
- large number of courses of action
● Optimization - finding values of - the problem belongs to a group and not
decision variables that minimize (or an individual
maximize) something such as cost - competing objectives
(or profit). - external groups are affected
● Objective function - the equation - problem owner and problem solver are
that minimizes (or maximizes) the not the same person
quantity of interest. - time limitations exist
● Constraints - limitations or
restrictions. 3. Structure the problem
● Optimal solution - values of the - Stating goals and objectives
decision variables at the minimum - Characterizing the possible
(or maximum) point. decisions
- Identifying any constraints or
Example: Prescriptive Pricing Model restrictions
A firm wishes to determine the best pricing 4. Analyze the problem
for one of its products in order to maximize
profit. Analytics plays a major role. Analysis
involves some sort of experimentation or
Analysts determined the following solution process, such as evaluating
predictive model: different scenarios, analyzing risks
Sales = -2.9485(price) + 3240.9 associated with various decision
Total revenue = (price)(sales) alternatives, finding a solution that meets
Cost = 10(Sales) + 5000 certain goals, or determining an optimal
solution.

5. Interpret results and make a decision

What do the results find by the model
mean for the application?
Models cannot capture every detail of the
real problem. Managers must understand
the limitations of models and their
underlying assumptions and often
incorporate judgement into making a
decision.
6. Implement the solution
Translate the results of the model back to
the real world.
Requires providing adequate resources,
motivating employees, eliminating
resistance to change, modifying
organizational policies, and
developing trust.

*CRISP slide - very last slide*