Everything to Know About Neuromatch Academy’s Computational Neuroscience Summer School
Welcome to the second article in the series where I share my experiences at Neurotechnology and neuroscience-related summer schools and events.
As has become the norm for me, I scroll LinkedIn every day, and this time, I happened to stumble upon a post on Neuromatch Academy — a 21 days long summer school on Computational Neuroscience (5th July — 26th July) taught by some of the most prominent professors and researchers in the field. With my recent fascination with all things Neurotechnology, Brain-Computer Interfaces related, I wanted to understand how neurons interact to give rise to millions of complex functions and how to use this electrical activity to decode our perception and cognition of the world around us.
They had a tentative course structure attached. It covered all the topics right from types of models to deep learning, dimensionality reduction, Markov models, control theory and reinforcement learning, and more — some topics I knew and others I was procrastinating on. I was excited beyond measure. It felt like I hit the jackpot. Some other features that caught my attention were that firstly, instead of watching videos and solving tutorials individually, everyone would be assigned groups and a TA who would conduct the tutorials and clarify the doubts much in a class like fashion, something almost lost in this virtual world of self-learning. Secondly, there was a dedicated project portion where students would collaborate to devise and work on a problem statement. The cherry on top of all these were the webinars that invited eminent researchers, entrepreneurs, and organizations in the field of Neurotechnology, brain-computer interfaces, and neuroscience to talk about their journeys, learnings, and career paths in this exciting and broad field. But two things scared me. How was I going to commit 8 hrs every day, and how will I understand such vast and complicated content in such a short period?
Nevertheless, I decided to apply to be an interactive student. Due to very high interest in the summer school but only limited TAs, everyone had to submit an application which was reviewed to select students who would most benefit from the program. The application was short and took me 20 mins to fill. It asked for my resume and past experiences in the field. Further, the application matched me with a TA in my time zone and language preferences. This made interactions very bilateral and conducive to learning. This was the first time that I was part of an event of such a large scale that accommodated different time zones and language barriers, making the program accessible to all students excited about computational neuroscience. If the organizers are reading this, I would like to thank you for being so thoughtful as I have grown up in India, where there are hundreds of languages, and sometimes communication is the biggest barrier to education.
I was overjoyed to receive an acceptance from the Neuromatch Academy. I started preparing for the summer school so that I could maximize my learning curve (a lot of new and challenging content to cover in a very short period). The 3 days of maths refreshers came to my rescue that revised concepts of linear algebra, differential equations and statistics, and probability. These concepts were the fundamental underlying concepts used to build the tutorials, and in hindsight, revision of these concepts helped grasp difficult topics. Moreover, there were short videos that introduced concepts and terms commonly used in neuroscience. I especially enjoyed watching these.
We were also introduced to our group members a week before the start. We all had very different technical backgrounds, some engineers, some studying medicine, and one person was studying economics, but we all were tied by our eagerness to learn more about computational neuroscience.
A typical day during the summer school started at 9 AM for me and ended at 6 PM. Yes, I know that is a lot. On some days, I worked after 6 PM on my project. But it never felt exhausting. I was always left excited about the next day. I fell into a rhythm. A rhythm that I fell in love with. Every day, we all went over the lecture videos, solved the problems together, and helped each other with our doubts. Some days the topics were known to me, but I loved learning about them from a neuroscience point of view, and on other days, the topics were completely new. When the topic was new, it was tough to follow the math as the course structure was very rigorous, but the tutorial was beautifully designed to methodically break down the tutorial into bite-size portions and explain each topic with interactive widgets that helped grasp the concept without going into the math. Further, there were intro and outro videos for every day’s topic that bound everything together and gave perspective. Personally, I enjoyed these the most as they gave the idea of the bigger picture and kept me curious and motivated.
As I mentioned, every day, we had a 3-hour long project slot along with our tutorial slot where we worked in a team of 3 to formulate and work on our problem statement. My team was multinational(Japan and India) as well as multi-disciplined( 2 engineers and 1 medical student). We decided to work on functional and organizational clustering of a human’s visual cortex using task-specific fMRI — more on this in another post. The entire 3 weeks were divided into writing an abstract for our project (which followed the what, how, and why modeling structure taught to us on day 1 of the summer school), working on the project, meeting with our project mentors, and project TA, and preparing a presentation to share our work with everyone on the last day. We were assigned 1 mentor who was an experienced researcher/professor in the specific topic area of our project and one project TA pursuing a Ph.D. in the field. They helped us formulate our problem statement, brainstormed ideas and results, and guided us throughout the project. Since none of our team members had any experience with fMRI data nor knowledge of the regions of the visual cortex, so much progress during such a short time would have been impossible without their guidance. Discussing ideas with them helped us with our project and taught us research practices followed globally by researchers in academia.
To further give us a larger picture and use cases of the topics we were learnings in computational neuroscience, there was a webinar every Friday where students could ask questions to the content creators. On the weekends, there were webinars where notable scientists and entrepreneurs like Thomas Reason, Philip Sabes, Konrad Körding, and many others shared their experiences and interacted with the audience. There were webinars on career design, graduate applications, mentor-mentee relationships, NIH scholarships, and more. Facebook Reality Labs, the main sponsors of the program, also gave a sneak peek into their exciting innovations, research trajectory, and future goals and discussed openings at the various teams.
To spice everything up, there were stand-up comedy, music, and logo-making competitions. We all voted to choose the winners, some of whom even performed during the closing ceremony. Another thing that I really appreciated about the program was the discord server. There was a channel for every possible topic, and the organizers, TAs, and content creators were super responsive. Even the webinar speakers were invited to interact with the students. The whole program and the discord server was a networking heaven.
To sum up, It is an excellent course for beginners who are curious to know more about CN as well as people already in the field. It demands a lot of time but is definitely worth it. Content is available for free online, so going back and revising is very easy. It’s a unique opportunity to make lifelong friends and collaborators. It was like being in school, in person, and I do not think it can get any better. Nostalgia is hitting me hard as I write the final few words.
Here are some tips to maximize your experience at the Computational Neuroscience summer school organized by Neuromatch Academy.
- Network — with students and mentors, and TAs. Take advantage of the Discord Server during and even after the program.
- Listen to as many people’s experiences as possible. It is a vast field and no single clear-cut path, so try finding a person whose interests most resembles yours and see their career growth and trajectory to understand the possibilities and opportunities.
- Refreshers are very important. Do not miss out on these if you want to maximize your learning curve during the following three weeks.
- Arrange for a week of break after the summer school to relax and return to some of the more challenging concepts. This revision will make all the difference.
- If you want to learn to code some concepts better, look into the code of helper functions.
- Check out their job board regularly
- Make friends — Nothing feels more like home than a like-minded community that shares your perspective and goals.
For further information, visit : https://academy.neuromatch.io/
Also feel free to check out their Deep Learning summer program.
