Instructor

Prof Veeraruna Kavitha (premidsem), Prof J. Venkateswaran (postmidsem)

Semester

Spring ‘20

Course Difficulty

3/5

Time Commitment Required

3 hours lectures + 2 hours self (per week)

Grading Policy and Statistics

The following are the statistics for the semester:
AA- 11
AB- 14
BB- 23
BC- 18
CC- 6
CD- 6
DD- 2
II- 7
Total-87

Attendance Policy

No compulsory attendance, but it is advisable to attend classes to ace the exams.

Pre-requisites

The knowledge of undergraduate probability will greatly ease the understanding of the course. This is because the course deals with analysis of multiple sets of data output from simulation softwares and concepts of distributions, confidence intervals, hypothesis tests etc will come in handy.

Evaluation Scheme

The evaluation was distributed in the following way:
Assignment 1 + Weekly Quizzes (premidsem): 15%
Midsem (completely in MATLAB/Scilab): 25%
Main Quiz 1 (postmidsem): 12%
Endsem (completely theoretical): 36%
End-term Project: 12%

Topics Covered in the Course

Premidsem: Basic probability distributions, conditional distributions, Poisson Point Processes, Markov Chains and characteristics

Postmidsem: Fundamentals of discrete event simulation, Input&Output Data Analysis, Comparing Alternatives, Design of Experiments, Simulation Optimization, ANOVA (briefly)

Teaching Style

The course content premidsem (taken by Prof Kavitha) was a complete mess, with she switching between topics as and when she desired. She briefly touched upon basic probability distributions, conditional distributions, Poisson Point Processes, Markov Chains and characteristics, and derived Little’s Law. The reference book for this was ‘Simulation’ by Ross. The book has a set of examples on discrete event simulation and one of them was asked to be coded in MATLAB/Scilab for the midsem.

The part taken by Prof JV was much better organized with proper slides and notes. There was no reference book needed, all notes were uploaded on Moodle. The prof makes sure that everyone understands, and his live lectures are very engaging. He is very understanding and allowed multiple deadline extensions (especially since this was the peak of the second wave).

Tutorials/Assignments/Projects

There was 1 assignment give premidsem, which was very easy. The end-term project is a bit demanding in terms of time. This is because it involves simulation on a software called Anylogic (which takes some time to get accustomed to). However the prof released the topics a bit late owing to which all groups had to struggle to get the project completed on time. Nonetheless the project is definitely rich in terms of learning outcomes.

Feedback on Exams

Weekly Quizzes (premidsem): They were moderately easy, those having completed one course in UG probability will have no problems acing these.

Test 1 (postmidsem): This was a completely theoretical test (on SAFE) and the checking was very lenient.

Midsem: This was a disaster, because the professor put no efforts on this. She just asked us to study the example set from ‘Simulation’ by Ross and asked one of this in the exam. Further the TAs did not even bother to check the code files we submitted (even after the prof promised this would happen), conducted one crib session which hardly 3 students attended, and did not reply back to grievance mails written to them.

Endsem: This was much better than midsem. It was theory based, and covered all topics from postmidsem + 10% premidsem. The grading was very lenient for this exam as well, and marks were not cut for minor mistakes.

Motivation for taking this course

This was the second course that I did as part of my IEOR minor.

Course Importance

This course is immensely useful in IEOR and for those dealing with massive simulation data outputs. I found this relevant for some future projects in my major (Chemical) as well.

How strongly would I recommend this course?

4/5 (simply because the postmidsem part was very engaging, and Prof JV will be taking the whole course this sem onwards)

When to take this course?

I took it in my 4th semester. The ideal semester is once you have completed an undergrad course in basic probability/data analysis.

Review By: Kshitij Sovanee