Blog


University of California, Los Angeles | Internship Interview

background

DTU Times interviewed Shreya Gupta, Software Engineering, Class of 2020, who secured a Research Internship at the Institute of Pure and Applied Mathematics, UCLA through the RIPS International Summer Internship program.

The general perception of research is that it is a long and trying road and does not offer a quick payoff, which people are often interested in. Considering this, what made you pick research over other options you might have had? What do you hope to achieve at the end of your journey?

It’s true that research is a long road, and often an exhausting one. You spend months getting to know a skill and to be eligible to apply it somewhere. If and when you finally get a project, you spend months investigating the topic and struggling around it, then you spend another couple of months trying to present your half-broken half-comprehensible work into a sophisticated paper and finally when you think you are done and submit it to your supervisor, you are told you might have to do the last few steps all over again. I will not sugar-coat it but it is one of the most exhaustive processes ever. It makes you feel emotions you never knew you had; it pushes you to extents you never knew you could go; it makes you do things you never imagined yourself doing; it pushes you absolutely out of your comfort zone and then traps you in itself. It will make you cry, and want to scratch your hair off. But when you finally see your name on the list of accepted papers, it squishes months of hard-work and sweat into that subtle smile. That is the beauty of research.

I never found purpose in anything I did. I could succeed in them, but never found the reason to come back to them. This is probably why I chose this field. In the end, I want to create and build things that improve the lives of people away from what has been considered the traditional 'development'. I want to build technology that reaches people who are underrepresented in the target audience of the technological advancements made so far. This includes people suffering from neurological diseases like autism who have hampered social well being because of their inability to recognise people. This includes early-stage cancer detection in women at a low cost. This includes finding solutions against rapid environmental changes. Essentially my hope is to apply tech in healthcare and environmental fields using CS and maths to achieve a broader aim.

What ignited your interest in your particular field of research? Having an interest in something and pursuing the field are two different things. What edged you to start your research?

In my freshman year, I was introduced to machine learning by my seniors in Pratibimb. I took the famous Coursera course and found it fascinating. I spent my sophomore year understanding and grasping it. I have always been a fan of maths so machine learning seemed the most suited amalgam of CS, tech and maths. I started writing emails and was accepted as a research intern in IITD. Doing the project independently was a mammoth task and made me realise the road is not going to be as simple. But the pain seemed worth it. There has been no looking back since then. I did self-undertaken projects, took courses and read books to further my understanding. It's too soon to say if this field is my calling but it is the closest I have gotten to being able to create applicative technology, so far.

What are some necessary steps any individual needs to take in order to be eligible for your particular scholarship/s? What stage is ideal to start preparing for them?

My advice would be to start early, do productive, and constructive research that serves some purpose. I started late and had to squeeze three years' work in a single year, so it was stressful. But the best time to start might be to explore various fields in the first year. 

  • Start working under a good professor from the summers of the first year or from the third semester. Keep coding (competitive) and do projects and research in machine learning. 
  • Try to cover any sub-fields (NLP, CV, AutoML, etc). 
  • Start applying for summer programs in the late third semester and in the fourth semester and make sure you work under a professor/research group, while doing competitive coding at the same time. 
  • It is also very important to maintain a good CGPA. For most programs a CGPA above 9 is desirable. 
  • For some that require you to contact a professor, an 8/8.5+ should also work given you have done quality research.
  • Be good at what you do. Start preparing in the third or fourth semester for whatever you want to do.
  • My accepting institution is the Institute of Pure and Applied Mathematics at UCLA, so maths matters to them.  I can not stress enough on maintaining a good CGPA!

What would help improve the chances of aspiring juniors looking to land a meritorious research internship? Are there any specific resources and skills you would like to recommend to aspirants?

Majorly the pointers from the previous answer. Understand what you are doing, read what has been done, be part of good communities (search on Meetup), and contribute to the field generously. Actively start adding content to your GitHub profile. Get your resume reviewed by tons of people.

Having research experience prior to applying for these programs helps. For resources, you can follow blogs like Google's AI research blog, KDNuggets, machinelearningmastery, and the likes (make Google your best friend). Follow good people, pages, hashtags and topics on LinkedIn to stay updated. I've personally never been a fan of extensive courses and would prefer doing a project over a course any day. It's the hard way but is always more constructive than a course.

Where does DTU lack in terms of research? What, according to you, can be improved here to keep pace with other leading research institutes?

I personally feel having better projects, will help DTU and it's students to raise the bar of research. Having real-life industrial and exciting projects helps gain insight into how to apply your skills to create technology that is usable and wholesome. Additionally, it also sets a deadline that makes work efficient.

I personally feel there has to be more dialogue and awareness about the research culture. In my interaction with fellow interns, I found their curriculum courses to be more diverse and in-depth. I think we can benefit from having more electives from the very beginning, with a wider variety of courses. Computer Science branches should ideally have more maths courses because maths is the fundamental of the majority of CS. There can be more seminars and workshops from eminent researchers and mentorship programs.

From my time spent researching in DTU, I have found the atmosphere to be bolstering in your field of choice, whatever that may be. There, however, exists a general lack of research atmosphere - the excitement to create something novel by an amalgam of things. Formulation of a research group that entails more like-minded people and challenges them to solve problems and engage in discussions might be a good starting point for the same. In general, more awareness about research needs to be prevalent from the freshman year itself. Under the current scenario, we have a crowd of lost people, like me, who want to do something but have no clue what or how.

Posted by Parangat Mittal

@thesciencestudent