SKANDA VAIDYANATH
Branch: Computer science
Batch: 2016 batch (4th year)
CGPA: 9.92
Field of interest: Reinforcement learning (a paradigm of ML). Currently interning at the Max Planck Institute for Informatics.
What is Reinforcement learning?
Reinforcement learning (RL) is a sub-paradigm of machine learning. Our college has courses that teach supervised and unsupervised learning, but this is a slightly different field. The reason I love RL is that it comes closest to the way humans think and behave. I am currently working on applying RL to information retrieval and natural language processing applications at the Max Planck Institute for Informatics.
How did you develop an interest in ML/RL?
Around my second year I started talking to a lot of professors and went on to take a course in machine learning. This was followed by Andrew Ng’s ML course on Coursera. Hours of classes taken by my seniors in ML also helped me get a clearer view and guided me. RL is something I started around my 3-1, I initially got into it because of its success at achieving human-level performance in video games and board games like chess and GO. Then an online NPTEL course on RL and textbooks I read before the summer of 3rd year got me started on this path. I came across several applications of RL in the real-world in robotics, healthcare and education. Two weeks into my third year summer internship I knew I wanted to do a doctoral degree in RL.
Any pointers on Projects, LORs and MS?
You need to be clear as soon as possible about what you want to do. Talking to professors helped a lot. Following my seniors helped as well. My senior went to Singapore for a project, so I wanted to go there too. In my PS 1 I was inclined towards the field of research (IGCAR) and I got a project of my own there. BE IN TOUCH WITH YOUR PROFS, regardless of whether it's for a project or not, and the guidance they give is crucial. LORs fall in place when you know professors and I got them from professors from foreign internships as well. Have a good relationship with them and talk to them outside class. Regarding MS (especially in the top 5), they are basically looking for reasons to reject you, so your CG, projects, thesis and LORs are crucial, and your resume must be the best it can be.
When did you start preparation for GRE?
I wrote GRE at the end of July 2019. Summer of third year, i.e. after 3-2 during my internship at USC Institute for Creative Technologies. The 10 days before GRE were hectic and I took a lot of tests online. Don't spend too much time on vocabulary, focus on reading comprehensions. It's easy to get 165 in math but not 170. This just comes with taking tests. I did not do any math outside the practice tests.
How did you manage clubs, departments and sports?
Do the right things at the right time. If you attend class during the day you have time to do other stuff later. You can't watch an entire sem’s worth of impartus a week before exams, it's bad timing. JUST ATTEND CLASS. If you go to class when you're supposed to go to classes you will be completely fine. You can always fit in gym, sports (I played tennis for 2hrs a day on average) and other activities in evenings when you go for classes regularly. I was also part of Depp, Dope and in IEEE in my 1st, 2nd and 3rd year, when I became vice chair.
Guidance to juniors:
1st year: It's initially hard. Try and concentrate on securing a high CG because it opens a lot of options. Some things just become much harder (example: foreign internships, good PS-1) if your CG is bad.
2nd year: Start seeing what's happening in the world of technology currently and look for what you are interested in. “The college gives a lot of importance to placements and hence students don't have exposure to research, and that's a problem! The only time you have exposure to research is in thesis or research internships. Search and decide what route you want to take. Choose between placement and research. Talk to faculty and research about it a lot. Make sure you know what “your ML/RL” is as soon as possible. This will help you build your profile/narrative towards graduate school applications as well."
3rd year: At the end of 3-1 you can start applying for research internships. Find some professor to work under (even if it's informally). This becomes useful for your resume and LORs. Summer internships can be taken up as well, but remember this is not through college (at least not the ones I applied to). This is the time to think if you want to pursue a thesis (important for MS) or PS2 (important for placement).
NOTE: Something needs to change about how research is viewed in our college. Concerned authorities should look into it as well. A lot of people go to PS because they don’t have a fair idea about research. At least explore this route. If anyone wants to talk to Skanda about research you can contact him on messenger.
Quarantine courses to check out: (Opinion 1: If college is giving you these courses, it’s better to take those instead of Coursera) (Opinion 2: NPTEL is better than Coursera, but it is very mathematical. )
Andrew NG Coursera ML
A-Z ML Udemy
A-Z Deep Learning Udemy
Python Data Science Bootcamp Udemy
CS 230 Stanford
CS 231N Stanford
Andrew Ng Deep Learning on Coursera
NPTEL Reinforcement Learning
NPTEL Deep Learning 1 and 2
“Make notes and don't watch it like it's Netflix.”
Disclaimer: The points given above are the views and steps taken by the individual. They are not fixed steps and guideline to base your college upon. Our hope is to inspire students so they can take the necessary steps here after. We hope you like it!
Written By: Manish Patil
Comments