GNR 652 – Machine Learning in Remote Sensing

Session –

2018-19

Instructor –

Biplab Banerjee

Course Structure:

1) 22.5% Midsem 2) 35% Endsem 3) 27.5% Course Project 4) 15% Coding Assignments

Course Content:

1) Supervised Learning 2) Unsupervised Learning 3) Semi-supervised Learning 4) Reinforcement Learning 5) Selected topics in deep learning (CNN, RNN etc)

Variety of current research techniques were covered under each of the above topics.

Course Project: Students were required to implement a research paper of their choice and demonstrate the results in the final presentation.

Prerequisite – Though there are no hard prerequisites, prior knowledge of probability and statistical mathematics would be helpful.

Feedback on the lectures – Attendance was not compulsory. The professor provided the slides for each lecture after the class. The pace of the lectures are quite slow, hence they are easy to follow and understand.

Feedback on Assignments/Tutorials – Tutorials were not held for the course but the TAs were always available to clear conceptual doubts. The assignments were of moderate difficulty and were based on implementing a certain algorithm or technique. The tests focused more on conceptual clarity rather than ability to solve tough mathematical problems.

Difficulty (on a scale of 1-5 with 5 being very tough) – 3

Textbooks/References – The professor didn’t follow any particular reference book. Detailed explanantions of almost all topics can be found by a quick Google search. The professor also provided relevant research papers for some topics.

Softwares used – MATLAB/Octave/Python

Reviewed by Anwesh Mohanty