Review by

Adit Akarsh, 2024(BTech. + CMInDS IDDP)

Course Offered In

Autumn 2022

Instructors

Prof. Udayan Ganguly

Prerequisites

Some coding experience in MATLAB/Python.

Difficulty

3/5 Content is not difficult to understand

Course Content The course covers the following topics-

1. Introduction: Describes the efficiency of the brain in contrast to digital devices used, highlights of the work undertaken by the reserach group.
2. Neurons: The functioning of neurons, with a brief focus on the electrochemistry associated with the spiking of neurons. Electrical models of neurons- Hodgkin-Huxley, Leaky Integrate and Fire, Izhikevich neuron.
3. Synapse: Biology of synapses, electrical model, potentiation and depression of synaptic connections, spike time-dependent plasticity.
4. Network of Neurons: Learning, a chemo-taxis neural circuit, a basic neural network, temporal vs rate coding for learning, Liquid State Machines for learning

Feedback on Lectures

The lecturs are heavy on student participation and discussions in small groups of students to understand concepts. The instructor is great at engaging the class. Questions in exam are similar to previous year papers and directly linked discussions in class. Attending lectures will greatly help.

Feedback on Evaluations

1. 15% Assignments (3 assignments in groups of 3): Coding assignments, fairly heavy. Difficult if you do not have prior experience in MATLAB or Python.
2. 60% Quizzes (4 exams of 1.5 hours each): Quizzes of equal weightage, covering only the portion covered after the previous quiz. Moderate difficulty, with 1-2 tough questions. Cover all the content as well as questions asked in red font in the slides.
3. 20% Course Project: In groups of 3, you can opt for a mentor (one of the TAs) and one of the projects they offer. While it is allotted on a FCFS basis, groups usually get their top preference. The project requires you to conduct a literature survey of the assigned topic for which the mentors provide some literature as a starting point, identify unsolved problems, suggest and implement novel improvements, and present the results.
4. 5% Class participation: Participate in discussion with the instructor and the class.

Study Material and References

1. Dale Purves- Neuroscience
2. JG Taylor- Computational Neuroscience

Follow-up Courses

None, but the research group is quite active. So, a project with the group can be considered if you like the field.

Final Takeaways

The course is a different in terms of the content covered; it does not align with any conventional EE field. The lectures of the course are quite unique; I have not seen any other course with this amount of student involvement in the lectures.

Grading Statistics: