Presenter Name	Digbalay Bose
Email-id:	digbalay@ee.iitb.ac.in
Topic:	Application of Sparsity based methods in image processing
Venue:	GG301
Date:	11/08/2015 at 4:00 PM
Pre-Requisite:	None
Abstract:	Sparsity based methods have generated lot of interests primarily in the domain of image processing. The session will be based on the basic idea of sparse representation, (along with some ideas about compressive sensing) and its application in specific domains like facial recognition, image denoising, super-resolution etc.

---

Presenter Name	Mohit Chhabra
Cluster	EE2
Email-id:	mohitchhabra@ee.iitb.ac.in
Topic:	Rubik’s cube:A group theoretic perspective
Venue:	GG-302
Date/Time:
Mentor:	Chayan Bhawal and Ameer Mulla
Session Chair:	Ameer Mulla
Pre-Requisite:	None
Abstract:	Group theory is a systematic way of studying symmetries. It can be used to reduce the computation required in many problems. This presentation gives an insight into the basics of Groups theory and how it can be used to solve Rubik’s cube from any possible orientation to home position in less than 52 moves.

Presenter Name	Apurva Joshi
Cluster	EE2
Email-id:	joshi@ee.iitb.ac.in
Topic:	Quadcopters: Making, Controlling, Flying
Venue:	GG-103
Date/Time:	23/09/2015 at 5:00 PM
Mentor:	Chayan Bhawal and Ameer Mulla
Session Chair:	Ameer Mulla
Pre-Requisite:	None
Abstract:	It's a bird! It's a plane! No no, it's a quadcopter. In this session we will look at the quad-rotor helicopter that has been developed at the Control and Computing Lab. We shall start by looking at the "making of a quadcopter" and then understand the control loops that make it fly. We shall look at the various modes of operation and then peek into the "swarm consensus" laws that we have been working on.

---

Presenter Name	Vinay Joshi
Cluster	EE2
Email-id:	vinay@ee.iitb.ac.in
Topic:	"Sensor Networks" - To solve real life problems
Venue:	GG-301
Date/Time:	09/10/2015 at 3:00 PM
Mentor:
Session Chair:	Chayan Bhawal
Pre-Requisite:	None
Abstract:	There is so much data around us and there are various real life problems we face. One can find solution to real life problems using desired data. Finding solution can be tricky depending upon a problem but one methodology that is largely being used nowadays is data analysis. Each and every problem can be solved if we have desired data with us. To obtain this data is not always a trivial task. In this session, speaker will take you through his experience of building reliable infrastructure using sensor networks for obtaining data and using this data to solve real life problems.

Presenter Name	Avishkar Palegar
Cluster	EE2
Email-id:	avishkar_palegar@yahoo.com
Topic:	Acceleration of Computational Electromagnetic PDE’s on GP-GPU using CUDA
Venue:	GG-301
Date/Time:	09/01/2015 at 4:00 PM
Mentor:	Deepak Anand
Session Chair:	Deepak Anand
Pre-Requisite:	None
Abstract:	Computational Electromagnetics problems are computationally costly because these problems involve solving Partial Differential Equations (PDE’s) which causes delay in execution time. Acceleration in solving these PDE’s will speed-up the process of solving the Computational Electromagnetics problems. The aim of the project is to do electromagnetic simulation using GP-GPU with the help of CUDA to accelerate partial Differential Equation (PDE’s) to 1000x times compare to CPU time. Development of Gauss Seidel/Jordan method solver of Partial Differential Equations (PDE’s) by Finite Difference Method (FDM) on CUDA enabled GPU using CUDA C/C++ language for acceleration compare to CPU time, is aimed in the simulation

Presenter Name	Ankit Vijay
Cluster	EE2
Email-id:	ankitvijay0190@gmail.com
Topic:	Finding the average number of jobs in a Random queue.
Venue:	GG-301
Date/Time:	05/2/2016 at 4:00 PM
Mentor:	Suhail Muhammad
Session Chair:	Suhail Muhammad
Pre-Requisite:	Basic Idea of Markov chain will be useful
Abstract:	The jobs in a random queue arrive according to a known distribution and are served (serviced) at different rates. Our objective is to quantify the average number of jobs, that is the stationary expectation, in the queue after the sufficiently long time in service. For this purpose, we build a Markov Chain model for different queues and discuss several of the existing methods to obtain their stationary expectation. We employ, Reinforcement learning algorithms, the Monte Carlo approach and the method of dynamic programming, and dwell into the drawbacks or the problems involved with each of them. Furthermore, we also explore the stochastic approximation scheme and it variants with function approximation for the infinite state space Markov model.
Presentation Slides:	Click Here.

Presenter Name	Anjan Kumar (Research Scholar, EE)
Cluster	EE2
Email-id:
Topic:	Mathematica for researchers: a hands-on session
Venue:	GG-303
Date/Time:	17/02/2016 2:00 PM
Mentor:
Session Chair:
Pre-Requisite:	Everyone is requested to bring laptops with Mathematica installed. Mathematica is available for download and installation at ftp://ftp.iitb.ac.in/IITB/Mathematica/Mathematica-10.0.2/. Installation instructions can be found at ftp://ftp.iitb.ac.in/IITB/Mathematica/README
Abstract:	Mathematica is a symbolic mathematical computation program, sometimes called a computer algebra program, used in many scientific, engineering, mathematical, and computing fields [Wiki]. In this session we shall look at the following topics: - Defining functions, manipulating expressions, - Solving equations, - Integration and differentiation, - Plotting - Interactive GUIs - Specific examples - Using palettes - Formatting This session would be useful for students across all specializations. Reference
Presentation Slides:	Click Here.

Presenter Name	Namita Behera, Post Doctoral Fellow (EE)
Cluster	EE2
Email-id:
Topic:	Nonlinear eigenvalue problem - A Review
Venue:	GG-302
Date/Time:	08/03/2016 at 3:35 PM
Mentor:
Session Chair:
Pre-Requisite:	None
Abstract:	The nonlinear eigenvalue problem is to find scalars λ ∈ C (eigenvalues) and nonzero vectors x and y (right and left eigenvectors) in \mathcal{C}^n$ satisfying F(λ)x = 0 and y ∗ F (λ) = 0, where F(λ) is an n × n nonlinear matrix function. In practice, the matrix elements are most often polynomial, rational or exponential functions of λ or a combination of these. These problems arise in many areas of science and engineering. They can be difficult to solve due to large problem size, and ill conditioning (that means the problem is very sensitive to perturbations and thus hard to solve accurately), or non availability of good numerical methods to solve them directly. In this seminar, the speaker aims to review the recent research ideas and directions in this area.
Presentation Slides:	Click Here.

Presenter Name	Shana Moothedath
Cluster	EE2
Email-id:	shanazuhara@gmail.com
Topic:	GATE: the math behind it!
Venue:	GG-302
Date/Time:	18/03/2016 at 3.45 PM
Mentor:
Session Chair:	Shivam Bhardwaj
Pre-Requisite:	None
Abstract:	Exam... Exam... Exam... All of us are evaluated through exams ... be it IIT-JEE, AIPMT, GATE or CAT. So it does matter to know, on what basis are these assessments being made. How is our eligibility being decided? Computerized Adaptive Test (CAT) is known as the best assessment technique for ranking students because of its optimal nature of exam conduction and evaluation. However, online adaptive exams require a pre-calibrated pool of questions. This set of questions need to be updated periodically. In this talk I would be discussing a maximum likelihood based alternating optimization algorithm proposed by us. Key words: Maximum likelihood, adaptive test, concave functions, alternating optimization.
Presentation Slides:	Click Here.

Presenter Name	Prof. S.D. Agashe (Formerly Prof.,Dept. of EE, IITB)
Cluster	EE4/EE5
Email-id:
Topic:	On Reading the Masters: Thought Processes of some eminent scientists of the past
Venue:	GG 002
Session Chair:
Date:	13/4/2016 at 5.00 PM
Pre-Requisite:	None
Abstract:	As students of Electrical Engineering, we have all studied Ampere's Law, Kirchhoff's Laws, Maxwell's Equations, and, almost certainly, Einstein's Special Theory of Relativity. But do we really know what they did, as could be found out from what they WROTE ? We may have some hunches but we could be wrong ! I will try to give some idea of what they wrote, partly by READING out some excerpts. If this is found interesting, we could later READ some of their writings. References: Behind the Beyond On the Solution of the Equations Obtained from the Investigation of the -Linear Distribution of Galvanic Currents, G. KIRCHHOFF. ON THE ELECTRODYNAMICS OF MOVING BODIES BY A. EINSTEIN