Computing for Data Sciences

Welcome to the 2017 edition of the course

Post Graduate Diploma in Business Analytics (PGDBA) - jointly offered by IIM Calcutta, IIT Kharagpur, and ISI Kolkata - aims to help shape the emerging profession of Business Analytics by delivering a cutting edge inter-disciplinary educational experience to its graduates.

Computing for Data Sciences (CDS), aka BAISI-4, is one of the five courses offered at ISI Kolkata during the First Semester of the PGDBA program. The Fall 2017 edition of the course -- CDS 2017 -- is taught by Sourav Sen Gupta from R C Bose Centre, Indian Statistical Institute, Kolkata.


Reach Sourav   :     sg.sourav@gmail.com   |     +91 94323 44852   |     Room 404, Deshmukh Building

  Information

Course : BAISI-4 (aka CDS)

Wednesday & Friday   @   11:00 - 13:00

Assignments = 30%     (group)
Mid-Sem Eval. = 20%     (individual)
End-Sem Eval. = 50%     (incl. project)


Format for evaluations to be decided.

  Lectures

We have time for about 26 two-hour lectures during Fall 2017 -- that's 52 hours in total! We will try to distribute this time carefully between Classroom Lectures (about 36-40 hours), Invited Talks (about 8 hours), and Interactive Sessions (about 4-8 hours) -- as required for the course. The schedule of the Classroom Lectures, and all relevant references and resources are posted as follows.

To prevent unwanted execution, the Python files (sometimes other code files too) linked below will be downloaded as *.py.txt. Please remove the .txt extension, and use the files normally.

Lec Date Topic Reference Resources
01 26-07-2017 Introduction to Computing   Demaine and Devadas : Lec. 1   Binary Search
02 28-07-2017 Introduction to Complexity   Cormen et al. (Algo) : Ch. 1, 2, 3   Asymptotic Notation
-- 28-07-2017 Introduction to Python   Python Class by Google   pythonIntro.py
03 02-08-2017 Introduction to Convergence   Demaine and Devadas : Lec. 11, 12   Newton's Method
04 04-08-2017 Finding Roots and Minima   Solomon (Numerical) : Ch. 8.1   Video 1   |   Video 2
05 09-08-2017 Linear Regression and Optimization   Andrew Ng : Lec. 2   Gradient Descent
06 10-08-2017 Linear Regression and Estimation   James et al. (ISLR) : Ch. 3.1
  Hastie and Tibshirani : Ch. 3 Lectures
  IntroLinReg.R
  Advertising.csv
07 16-08-2017 Linear Regression Fundamentals   James et al. (ISLR) : Ch. 3.1, 3.2
  Hastie and Tibshirani : Ch. 3 Lectures
  LinRegBasics.R
  Advertising.csv
08 17-08-2017 Linear Regression and Projection   Gilbert Strang : Lec. 10
  Andrew Ng : Lec. 2
  The Sub-Spaces
  Normal Equation

  Books and Compilations

Textbooks and Recommended References

There is no single textbook for CDS. The following books and compiled lecture notes treat, quite comprehensively, the topics that CDS broadly tries to cover -- highly recommended.

  • Chambers

    Software for Data Analysis: Programming with R

    John Chambers Springer (Second Printing), 2009
  • CormenEtal

    Introduction to Algorithms

    Thomas H. Cormen, Charles E. Leiserson, Ronald L. Rivest and Clifford Stein The MIT Press (Third Edition), 2009
  • Solomon

    Numerical Algorithms

    Justin Solomon Available online at this link
  • ISL

    An Introduction to Statistical Learning

    Gareth James, Daniela Witten, Trevor Hastie and Robert Tibshirani Available online at this link
  • ESL

    The Elements of Statistical Learning

    Trevor Hastie, Robert Tibshirani and Jerome Friedman Available online at this link
  • MMD

    Mining of Massive Datasets

    Jure Leskovec, Anand Rajaraman and Jeff Ullman Available online at this link

The books listed above are the basic references for CDS. During the course, we will refer to these books from time-to-time, as and when required, but we may not see through any of these books cover-to-cover.

  Lecture Videos and Notes

Adds a completely new dimension to Reading

There are several lecture notes and video lectures that perfectly complement the material that CDS plans to cover. The students are encouraged to follow these amazing resources.


The courses and video lectures listed above are closely related to the topics covered in CDS. During the course, we will refer to these lectures from time-to-time, as and when required, but we may not need to go through any of these courses end-to-end. In addition to the above, please feel free to explore Coursera.

  R Programming Resources

The main tool for computing used during CDS is R. The following resources may help you in getting yourself acquainted with the basics of R. Get comfortable, and get your hands dirty!

  Python Programming Resources

The most versatile language for Scientific Computation is Python. The following resources may help you in getting yourself acquainted with the basics of Python. Get used to it!

  Assignments

Worth 30% of the total grade

Assignments will be posted on a regular basis. Each group has to submit 6 assignments during the course of the semester, out of which top 5 will be counted towards the grade.

Assignment Posted Deadline Remarks
  Assignment 0 28-07-2017 09-08-2017 Not graded
  Assignment 1 07-08-2017 15-08-2017 Individual
Assignment 2 17-08-2017 25-08-2017 To be posted
Assignment 3 28-08-2017 05-09-2017 To be posted
Assignment 4 12-10-2017 20-10-2017 To be posted
Assignment 5 23-10-2017 31-10-2017 To be posted
Assignment 6 02-11-2017 10-11-2017 To be posted

  Projects

Worth 30% of the total grade

About 30% weightage (out of 50%) will be reserved in the End-of-Semeseter evaluation for the Term Project. Each group is supposed to deliver a Project Presentation (30 mins per group), including a Q&A session (5 mins per group), and a Project Report (theory/code).

Each group is at the liberty of choosing the topic for their Term Project. However, the term project chosen by each group must be approved by Sourav before they may be executed. The last date for finalizing the topics for the term projects is Thursday, 7 September 2017.


One may take part in a competition on Kaggle, DrivenData, CrowdAnalytix, InnoCentive, etc, or explore the open datasets made available by UNICEF, WHO, AWS, Google, EU, and US Govt.

Term Projects may also be inspired by earlier Stanford CS229 projects available at 2016, 2016 Spring, 2015, 2014, 2013, 2012, 2011, 2010, 2009, 2008, 2007, 2006, 2005, 2004, or by the diverse list of Business Case Studies posted by the Sloan School, MIT.

Potential topics for Term Projects may also include substantial extension (upon mutual discussion with Sourav) of previous years' projects available at CDS 2016, CDS 2015.