Business Analytics Course

Module 1: Business Analytics (Internal)

Basics of Big Data
Analytics Maturity
What is Data Science
Introduction to Hadoop
Analytical Tools: R vs SAS
What is Machine Learning
Module 2: Tools (Internal)
Introduction To R
o Data Import Techniques
o Analyze and Processing of Data
o Plotting function in R
o Data Sub setting: Indexing
o Functions in R
o Integrating R with Hadoop
Introduction To Tableau
Introduction To Python
SQL Queries
Module 3: Machine Learning (Internal)
Basics of ML: Supervised vs Unsupervised
Use Cases:
o Forecasting
o Classification
o Clustering
o Association Detection
o Anomaly/Outlier Detection
o Dimensionality Reduction
o NLP (Natural Language Processing)
o Image Detection & Deep Learning
Algorithm Basics:
o Regression (Linear & Logistic)
o Bayesian Methods (Nave)
o Nearest Neighbour
o SVM (Support Vector Machine)
o Neural Network & Preceptors
o Decision Tree, Forest & Jungle
o Artificial Neural Network
o Principal Component Analysis
o Time Series Modelling
Algorithm Cheat Sheet (Which Algorithm To Use When)
Module 4: Case Studies (External)

Prediction
o Demand Forecasting
o Price Prediction
Clustering o Market & Customer Segmentation
o Event/Activity based Segmentation
o Topic Modelling

Classification
o SPAM Detection
o Churn Analytics

Anomaly Detection
o Credit Risk Scoring
o Fraud Detection
o High Speed Arbitrage Trading

Association Detectiono Market Basket Analysis

o Next Best Offer Analysis
o Recommendation Engine (Products, Friend , Playlist)
o Cross Sell Model
o Customer Lifetime Value Analysis
o Sentiment Analysis

Dimensionality Reduction
o Genre classification
o Quant Trading

Neural Network
o Social Graph Analysis
o Quant Trading

Natural Language Processing

o Social Graph Analysis
o Topic Modeling
o Text Analysis
o Sentiment Analysis

Other
o
o
o
o
o
o
o
o
o

Cross Channel Analysis

Channel Optimization
Campaign management and Optimization
Discount Targeting
Event/Behaviour Based Targeting
Call Detail Record Analysis
Network Performance Optimization
Dynamic Pricing
Loyalty Program

Module 5: Data Visualization ((Internal)

Introduction
Data Visualization Process
Working with Single Data and Multiple Data Sources
Using Calculations in Tableau
Comparing Measures Against a Goal
Tableau Geo coding, Advanced Mapping
Showing Distributions of Data
Statistics and Forecasting
Dashboard Best Practices
Case Study

Module 6: Big Data & Hadoop (Internal)

Introduction To Hadoop Technology

Basics of Hadoop
HDFS and Cluster Environment
Map Reduce basics, Types and Formats
HIVE/PIG/HBASE/SPARK etc

Case Study 1: Customer

Segmentation
Every customer is different and so is their behaviour. To understand the vast and
diverse universe of the customer base, we need to segment the customers based
on their characteristics. These characteristics could be based on the
requirements, preferences, lifestyles or life stages.

Task 1: Design a model which can segment customer

into different groups
Data: Customer Orders Data_RFM.CSV
High level flow:
1. Calculate R,F and M parameters
2. Calculate Breadth & Tenure
3. Mosiac Plot for R vs F, M vs F, M vs R, Create a balloon Plot and show
different segment
a. Life Stage ( New, Active, Lapsed, Lost)
b. Value ( Gold, Silver, Bronze)
c. Combined as (High Value, Repeat, New, One-time, Lapsed, &
Lost)
4. Apply K-Mean clustering algorithm on these parameters to group
similar customers
5. Apply classification algorithms such as Logistic Regression and
Decision Trees to predict future customer behaviour.

Task 2: Also try to answer below questions

How is each customer segment different to other?
Which customers are likely to churn?
Which set of customers have potential to spend more?
Which products or services entice which customer group?
How seasonality does affect the customer behaviour?
What is the intersection between the customer loyalties with their
life stages?

RFM Analysis (Balloon Plot)

Task 3: Create a presentation which is DEMO ready with

below information

Problem Objective:
Approach:
o Data Description:
o Algorithm Used:
o Language/Tool Used:
Output:
Challenges:

Reference:
1. https://ds4ci.files.wordpress.com/2013/09/user08_jimp_custseg_revnov08.
pdf
2. http://inseaddataanalytics.github.io/INSEADAnalytics/Report_s45.html
3. http://analyzecore.com/2015/02/16/customer-segmentation-lifecycle-gridswith-r/

Case Study 2: Market Basket

Analysis
Market Basket Analysis (MBA), also known as affinity analysis, is a data mining
technique used to understand the purchase pattern of the shoppers. This
analysis unveils the hidden association between the products or the product
groups. This analysis is highly recommended for understanding cross-selling
opportunities in the associated categories as well as up-selling opportunities
within categories. Advanced Market Basket Analysis provides an excellent way to
get to know the customer and understand the different behaviours.

Task 1: Design a model which can segment customer into

different groups and try to answer below additional
questions

Summary of Purchase behaviour

Develop a cross promotional programs
How this analysis can help redesigning store layout/design
Prepare a discount plans / promotions

Data:
http://archive.ics.uci.edu/ml/datasets/Online+Retail
Grocery.csv

Task 2: Create a presentation which is DEMO ready with

below information
Problem Objective:
Approach:
Data Description:
Algorithm Used:
Language/Tool Used:
Output:
Challenges:
Reference: https://gallery.cortanaintelligence.com/Experiment/Frequentlybought-together-market-basket-analyses-using-ARULES-1

Case Study 3: Customer Churn

Analytics
Today, a retailer needs to keep track of the rate of attrition in their customer
base. Acquiring new customers is far more costly than retaining existing
customer as retaining customers maintains volume and revenues at lower cost,
increases customer lifetime value, and should create loyalty from limited
experience and habit. Customer Churn analysis help organizations identify key
factors driving churn and to design customer churn prevention and win back
strategies to improve customer retention, loyalty and Life time value.

Task 1: Create a Churn Analytics model which predict

the Churn and try to answer below question

What are my key Churn drivers?

How many customers am I going to lose by next month?

How can I develop a customer win back strategy?

Data:
UserInfoData.CSV : User Information Data
UserActivityData : User Activity Data

High Level Flow:

Reference: https://gallery.cortanaanalytics.com/Collection/Retail-Customer-ChurnPrediction-Template-1

Case Study 3: Pricing

Optimization
High competition, consumer price sensitivity and pressure on margins are the
key challenges that a retailer faces every day. Pricing optimization has been
shown time after time to increase sales and margins; it is one of the most direct
routes between analytics and the bottom line. It is designed to enable retailers to
optimize base prices to achieve their sales, volume, profit and price image
objectives for regular, everyday items.
Data: Point-of-sale information for seasonal and non-seasonal data
Create a model which can forecast a set of demand curves for particular SKUs in
particular stores or clusters. The demand curve identifies the products that are
the most and least price-sensitive.

Also try to answer these additional questions

From these predicted

recommended pricing.

How can I price my brands to increase the overall portfolio growth?

What is the impact of competitive price on my brands?

Which products or brands are highly price sensitive?

What are the right price points to maximize the sales?

What is the right mix of Price discount and marketing activity?

demand

curves

how

to

determine

optimal

Case Study 4: Marketing Mix

The old complaint that, Half of the money I spend on advertising is wasted; the
trouble
is
I
dont
know which half no longer has to be true.

Through marketing mix modelling, it is determined where and how much the
retailers
should
spend
their marketing resources. We use econometric models, along with the deep
understanding
of
your
market and industry, to measure the performance of your marketing mix. Our
optimization
techniques are flexible and are specific to the business challenge; hence
maximize
sales
and
produces highest return for your marketing investments.

Key Features:

Measuring Effectiveness Determine which media vehicles are most effective

at
driving
sales, profits, share and consumer segments
Understanding sensitivity Understanding whether the macroeconomic,
demographic,
competitive, seasonal, weather and operational factors impact on sales and
shares
ROI Calculation ROI is calculated for each advertising and promotional
campaign,
including halo effects and cannibalization on related brands
Recommendations The results of our marketing mix modeling tells us whether
a
marketing vehicle needs to continue or discontinued and how to reallocate the
marketing
spends for maximizing marketing ROI.

Case Study 5: Price Prediction

Tubes can vary across a number of dimensions, including base materials, number
of bends, bend radius, bolt patterns, and end types.

Currently, Caterpillar relies on a variety of suppliers to manufacture these tube

assemblies, each having their own unique pricing model. This competition
provides detailed tube, component, and annual volume datasets, and challenges
you to predict the price a supplier will quote for a given tube assembly.

Data:
Sample_submission.csv
Data.csv

The dataset is comprised of a large number of relational tables that describe the
physical properties of tube assemblies. You are challenged to combine the
characteristics of each tube assembly with supplier pricing dynamics in order to
forecast a quote price for each tube. The quote price is labeled as cost in the
data.
High Level Flow:

Reference: https://www.kaggle.com/c/caterpillar-tube-pricing