Sankhya Analytical Research

Decision making with statistical intelligence

Diploma in Data Science

Private and confidential 1

 Course Features

• Course design based on 12+ years industry experience

• Exposure to real world analytics

• Hands-on working on R and Python

• Business case examples from domains such as BFSI, FMCG,

Retail, HR, Telecom, Pharmaceutical, etc.

• Course guided by highly skilled and experienced trainers

2
Training Program Framework

3
 Focus Areas

A data scientist represents an evolution from the data analyst role.

A strong foundation in Statistics remains a formal requirement.

Analytics skills Domain knowledge

A data scientist’s
skill set

Programming skills Communication skills

4
 Course Structure

Assignments
60 hours
Classroom Training

40 hours

Total
Project Work 200
80 hours hours

Evaluation
20 hours

5
 Module 0: Introduction

 Introduction to Data Science

 Business Analytics success stories

 Exposure to tools like R, RStudio, MS Excel,

MySQL

 Scope and opportunities

 Fundamentals of Statistics

6
 Module 1: Exploratory Data Analysis

 Data Management
– Import, sort, merge,  Critical for successful analytics
aggregate, subset, derive implementation
– Introduction to MySql

 Descriptive Statistics  Good data management helps to

– Central tendency – assess quality of data
– Variation – improve the quality of data
– Shape – make data analysis ready

 Visualisation  Provides data insights

– Bar charts/histograms
– Box-Whiskers plot
– Contour plot  Guides towards business research
– Motion chart problem solution using advanced
analytics

4 hrs 7
 Module 2: Statistical Interference

 Distribution Theory and

Hypothesis Testing  Powerful tool for testing
– Discrete distributions researcher’s claim in the planned
– Continuous distributions experiment
– Parametric tests
– Non-parametric tests  Wide application in clinical, market
– Analysis of Variance and social research
– Analysis of Covariance
 Marketing campaigns can be
designed and tested before full
fledged implementation

4 hrs 8
 Module 3: Predictive Modelling - Fundamentals

 Basics of modelling
– Modeling framework  Growing area in Risk Management
– Best practices & Marketing

 Multiple Linear Regression  Cross selling/up-selling can be

– Mathematical model done scientifically
– Validating assumptions
– Residual analysis  Financial institutions can predict
– Multicollinearity problem ‘bad’ customers
– Out of sample validation

 Huge scope in e-commerce

business

8 hrs 9
 Module 4: Predictive Modelling - Advanced

 Categorical Response
Variable  Categorical response variables are
– Binary logistic regression frequently incorporated in real
– Multinomial logistic world scenarios
regression
– Ordinal logistic regression
 Most widely used class of
– Poisson regression
(modelling count
predictive modelling
response variable)
– Cox regression  Response to offer – Yes/No

 Brand preference –
iPhone/Samsung/Sony

6 hrs 10
 Module 5: Time Series Analysis
 Time Series Modelling
– AR models  Set of models of forecasting sales,
– MA models financial indices, economy indices
– ARIMA
– ARCH  Inflation rate ,GDP are predicted
– GARCH using time series modelling
– Time Series regression
– Exponential smoothing
 Nifty/Sensex future values can be
estimated

 Complex financial models are

developed using ARCH & GARCH

4 hrs 11
 Module 6: Unsupervised Multivariate Methods

 Segmentation and data

reduction  Provide exploratory segments of
– k-means clustering customer, stores & agents
(algorithm and selection
of best cluster solution)
 PCA/Factor Analysis are powerful
– Principal component
analysis
techniques for dimension reduction
and scoring models
– Principal component
regression
– Factor analysis  PCA is used to resolve
– Multidimensional scaling multicollinearity problem in
regression models

4 hrs 12
 Module 7: Machine Learning
 Machine Learning
Algorithms  New generation algorithms
– Naïve Bayes
– Support vector machines  Multiple methods can be used to
– Decision tree decide best predictive model
– Random Forest algorithm
– Neural networks
– Association rules
 Discover hidden pattern which
may not be revealed by classical
– Introduction to KNIME
and RATTLE methods

6 hrs 13
 Module 8: Text Mining

 Data Management using

 Volume & velocity of the data is
Python
humungous
 Exploring Pandas library
 Connecting R with Python  Platform for analytics
to perform Statistical implementation in cloud
Analysis environment
 Sentiment analysis of
Facebook and Twitter data  Combines unstructured data with
(Text Mining) structured data

4 hrs 14
 Assessment
Assignments

Individual presentation 25
and Viva marks
25 marks

Project work Total

300 Module tests
50 marks
marks 200 marks

15
 Eligibility Criteria

• Must have studied Mathematics at 10+2 level (or during

Graduation years)

• Final Year Graduation/Graduate/Post Garduate/Ph.D students

from any background

• Currently Pursuing Post Graduation/Ph.D from any background

• Working Professionals from any background

Course Duration: 6 months ( can be completed in 3 months)

16
 About Sankhya

• We are a knowledge process outsourcing firm based in Mumbai and

London, providing training and consulting in the field of Advanced Analytics

• Our vision is to empower our Data Science students to leverage the power
of data and make data-driven decisions in business and research

• Our promoter team has a cumulative experience of 40+ years

• We offer a high level of expertise in the subjects of Statistics coupled with

efficiency in R, SAS and SPSS

• Our clientele include reputed organisations across industry verticals: BFSI,

FMCG and Pharmaceutical

17
 Sankhya’s Legacy

50+ clients

PGDDS at
St. Xavier’s 100+ projects
College Mumbai
Sankhya

50+ corporate
12 research
training
papers
programs

18
Our Esteemed Clientele

19
 Vinayak Deshpande
• M.Sc. Statistics, Mumbai University (1997),First Rank
• Professor-Ramnarain Ruia College (1997-2004)

• Founder, Managing Director: Sankhya Analytical Research Pvt. Ltd. (2004)

• Director: Medicounts Life Sciences Pvt Ltd(2013)
• Principal Advisor: Direxions Marketing Solutions Pvt. Ltd. (2013)

• Provided services to reputed clients in India, USA and Korea in BFSI, Retail, FMCG
and Pharma domains

• Published papers on biostatistics, marketing analytics ,econometrics Member- Board

of Studies, St.Xaviers College
Worked as a member of Independent Ethics Committee for 2 clinical trials

• Recent research papers:

1) Evaluating Principal Component Regression in case of severe
Multicollinearity.
2) Comparison of predictive modeling methods: Naïve Baye’s, Logistic 20
Regression and Support Vector Machines
 Sheetal Joshi
• M.A. Economics, Mumbai University (2004)

• Director: Sankhya Analytical Research Pvt. Ltd. (2005-2010, 2013 onwards)

Instrumental in providing data driven solutions in Marketing , Market Research, Risk

Management and Clinical Research for clients across BFSI sector in India and USA

Highly experienced in designing and delivering corporate training programs in

SAS/R/Business Analytics

Customized clinical data management tool Open Clinica for global clinical trial

• Started Market Management department at Bajaj Allianz General Insurance

Responsible for Initiatives in Customer Lifetime Value Management , Brand

Performance Measurement and Net Promoter Score
Developed platform for Lead Management System, Created single customer view 21
and campaign management (2011-12)
Sankhya Analytical Research

www.SankhyaAnalytics.com

Private and Confidential 22