BUSINESS ANALYTICS

Instructor: Theu DINH

1/28
1 / 28
Course Structure

1 Session 1: Introduction

2 Session 2: Descriptive Analytics

3 Session 3: Predictive Analytics

4 Session 4: Prescriptive Analytics

2/28
2 / 28
Course Objectives and Learning Outcomes
Objectives
This course aims to introduce students to key concepts in business
analytics, with a special emphasis on common areas of application
Help students understand what a Business Analyst is, what they do,
and how they do it
Guide students how to make and communicate data insights
It also explores the links between the behavioural (decision science)
perspective and the management science perspective
Master data science tools: Excel, R, Data Visualization Data
Storytelling
Learning Outcomes
Critically evaluate decision science and business analytics perspectives
on organisational decision support and identify and explore their
complementarity.
Explore and comment on the emerging role of ‘big data’.
Display familiarity with major areas of application of business
analytics. 3/28
3 / 28
Session 1 / DEFINITION OF BUSINESS ANALYTICS

Definition
The process of using quantitative methods
HBS to derive meaning from data to make
informed business decisions.
Taking in and analysing historical business data
to identify trends, patterns, and root causes
ORACLE
before making data-driven business decisions
based on those insights
A set of disciplines and technologies for solving
business problems using data analysis,
TechTarget
statistical models and other quantitative
methods
4/28
4 / 28
Session 1 / WHY IS BUSINESS ANALYTICS
IMPORTANT?

Improved decision-making
Provide insights into customer behavior and habits
Identify business needs
A culture driven by data
Minimizes risks and reduce
Improved performance
Improved strategic planning
Increase day-to-day operational efficiency

5/28
5 / 28
Session 1 / BUSINESS ANALYTICS TOOLS

Database queries and analysis

Dashboards to report key performance measures
Data visualization
Statistical methods
Spreadsheets and predictive models
Scenario and “what-if” analyses
Simulation
Forecasting
Data and text mining
Optimization
Social media, Web, and text analytics

6/28
6 / 28
Session 1 / BUT HOW TO BUSINESS ANALYTICS?

Figure: Four Levels of Data Business Analytics

7/28
7 / 28
Session 1 / DESCRIPTIVE ANALYTICS

Descriptive Statistics
Data Visualization
Probability Distributions and Data Modeling
Statistical Inference

8/28
8 / 28
Session 1 / PREDICTIVE ANALYTICS

Regression Analysis
Forecasting Techniques
Predictive Data Mining
Spreadsheet Modeling and Analysis
Monte Carlo Simulation and Risk Analysis

9/28
9 / 28
Session 1 / PRESCRIPTIVE ANALYTICS

Optimization
Decision Analysis

10/28
10 / 28
Session 1 / BUSINESS INTELLIGENCE (BI) VS.
BUSINESS ANALYTICS (BA)?
BI and BA use data to help businesses make better decisions, but
they do it differently

11/28
Figure: BI and BA 11 / 28
Session 1 / BUSINESS INTELLIGENCE (BI) VS.
BUSINESS ANALYTICS (BA)? (cont.)

Figure: BI and BA difference

12/28
12 / 28
Session 1 / DIKW Pyramid

Figure: Model DIKW

13/28
13 / 28
Session 1 / The Challenging Journey of Data Business
Analytics

Figure: Data Analytics Marathon

14/28
14 / 28
Session 1 / The Role of Data in the Business World

Figure: Role of data in business

“The world’s most valuable resource is no longer oil, but data”

(The Economist, 2017)
15/28
15 / 28
Session 1 / The Role of Data in the Business World (cont.)

Data is one of the most valuable assets of any organization

Data is Power in many various industries
More accurate prediction and better-informed decision-making
An organization will have more possibilities to improve their daily
productivity if they adopt a data-driven culture
How to transform all data into meaningful business
outcomes?
Categorize data
Synthesize insights from data categorization
Create an effective way to share those insights

16/28
16 / 28
Session 1 / Big Data

Figure: 5V’s of big data (Source: CloudDuggu.com)

17/28
17 / 28
Session 1 / Big data vs Traditional data - Volume

Trad. data Big Data

Small to medium-sized datasets Extremely large datasets
Volume of data Its volume ranges from Gigabytes to Terabytes From Petabytes to Zettabytes or Exabytes
Using traditional data processing technologies Cannot use traditional technologies

Figure: Volume growth of data (Source: CloudDuggu.com)

18/28
18 / 28
Session 1 / Big data vs Traditional data - Variety
Trad. data Big Data
Structured data Structured, unstructured, or semi-structured data
Derives from enterprise resource planning,
Variety of data Derives from a broader range of enterprise
customer relationship management, online
and non-enterprise-level data
transactions and other enterprise-level data

Figure: Variety of big data (Source: CloudDuggu.com)

19/28
19 / 28
Session 1 / Big data vs Traditional data - Velocity

Trad. data Big Data

Static Dynamic
Velocity of data Is generated per hour or per day or more Is generated more frequently mainly per seconds
Data model is strict schema based Data model is a flat schema based

Figure: Velocity of big data (Source: CloudDuggu.com)

20/28
20 / 28
Session 1 / Big data vs Traditional data - Veracity
Big data can be more messy, noisy, and error-prone than traditional
data. These make it difficult to control the quality and accuracy of
big data than traditional data

Figure: Veracity of big data (Source: CloudDuggu.com)

21/28
21 / 28
Session 1 / Big data vs Traditional data - Value
Big data can provide higher potential value into customer behavior,
market trends, and other business-critical information than traditional
data.

Figure: Value of big data (Source: CloudDuggu.com)

22/28
22 / 28
Session 1 / Similarities between Big data and Traditional
data

Data Analysis: Both traditional and big data require some form of
analysis to derive insights and knowledge from the data
Data Storage: In both traditional and big data environments, data
needs to be stored and managed effectively
Business Value: Both traditional and big data can provide
significant value to organizations
Data Security: Data security is a critical consideration in both
traditional and big data environments
Data Quality: The quality of data is essential in both traditional and
big data environments. Accurate and reliable data is necessary for
making informed business decisions

23/28
23 / 28
Session 1 / Questions for discussion

Q1: Why traditional data still accounts for the majority of the world’s
data?
Q2: The rise and benefits of big data can make traditional data is
going away or not?

24/28
24 / 28
Session 2

25/28
25 / 28
Session 3

26/28
26 / 28
Session 4

27/28
27 / 28