From English Literature to Data Science

How did I land a job in data science?

I was thinking about posting here on a more consistent basis, and what’s better than starting with my own story? I made the transition that many want to, and landed a job in data science coming from a non-technical background (no STEM degree, no social science degree). It wasn’t very straightforward: I moved in a zig-zag fashion, so try to bear with me.

Photo by Tyler Nix on Unsplash

Developing Interest In Science

It was my first year in university and I wasn’t interested in anything other than basketball. I was studying English Literature only because my English was better than my peers’ back in the day, and I was able to get into a university with it. Anyway, during the semester break of that year I went to Lebanon, Beirut for human biomechanics training (during those days I was training basketball players ). While I was there, the effect of prolonged stress was mentioned a couple of times and it sounded interesting. When I came back home, I wanted to explore more and watched Dr. Robert Sapolsky’s TED Talk titled The Biology of Our Best and Worst Selves. I was so impressed, I remember going like “I wonder how he (Sapolsky) sees the world.” Luckily, he had his whole class on Stanford’s YouTube: Human Behavioral Biology. Well, I couldn’t keep up with it due to my lack of background (didn’t know anything related to science back then, this includes biology and/or neurobiology/psychology). I decided to visit the psychology department.

Getting Into Psychology Lab

Whole purpose of me visiting the psychology department was about a lecture again which I can’t seem to remember now but it was related to evolution of belief systems. I went in by the first door that I found and met with Dr. Bahcekapili and he told me to introduce myself to his colleague (his student then) next door, Dr. Yilmaz. From then on, my interests shifted towards his: Intersection between morality, politics, religion, and decision making. I started to take his social psychology class unofficially and started to stop by his door more often to ask about things that I read: behavioral biology to social psychology, and evolutionary psychology. At one point, he asked me to be part of the lab that he was about to found and I said yes, ended up in MINT Lab.

Learning Statistics & Mathematics

Before the start of the lab, I enrolled in Science of Religion in edX to be sure that I was not missing important findings in the field. It was enjoyable because the information was presented in a very graspable way. But with the lab, once we started reading research papers, I realized I don’t know anything about statistics hence I was not able to evaluate how convincing the research in front of me is. Being a literature student, I had to find something that assumed zero prior knowledge. I tried many books but OpenIntro Statistics and Learning Stats with JASP helped me a ton and they were enough for an undergraduate level. I started to realize how much I enjoy studying statistics and wanted to improve myself on the matter, wondered if mathematics could help with statistics.

There was someone that I got to know by playing Hearthstone (yes, I enjoy card games): Dr. Basar Coskunoglu. I started studying mathematics with him. He was patient enough to start from very basics (I didn’t know anything, literally, so we had to start from functions, inequality systems etc.) with me. We made it all the way to the calculus and linear algebra. We finished at a point where it was possible for me to go through books by myself, which was what I wanted in the first place. To this day, I still study linear algebra from different sources (mainly from my notes, Mathematics for Machine Learning and Gilbert Strang’s books) almost weekly since I enjoy it very much.

From Lab to Data Science Internship

In lab and academia (social sciences), researchers mainly use statistical softwares such as JASP, Jamovi, SPSS. R programming is also used but during those times I started to lose a bit interest in academia due to different reasons. So, I wanted to learn Python. I attended a boot camp (I strongly advice against any data science boot camp now, every piece of material they provide is available online and probably in a better way) which mainly helped me with programming skills along with industry-based cases rather than theoretical part since I was already done with the classic (up to deep learning): Introduction to Statistical Learning (Python version is available now).

At the time I was still an undergraduate and was on a scholarship from TUBITAK (The Scientific and Technological Research Council of Turkey) for taking a part in a research project but it was coming to an end. So, timing was really nice to try my chances outside the academia. I wanted a part-time job or a long-term internship since I was still at school and did not want to burn out. Anyway, I started doing data science projects and included them on my CV under the “projects” part. I applied to different jobs but during the interviews I realized that many did not have a data science team and it scared me a little since it was going to be my first job, I felt like I needed a group of people that have some experience. Although I wasn’t sure how I would fit in a bank, I applied to DenizBank’s long term internship and after some steps, I was accepted as a data science intern.

Transition to Full-Time

I felt like I did OK during my internship, and it looks like this feeling was mutual since they wanted to keep me (also, it makes sense to keep someone after 6 months rather than hiring a new person). Everything went smoothly at school, and I graduated. Additionally, although I spent six months there coding etc., it was still required for me to take the junior data scientist test which included SQL, Python/R and probability and statistics questions. I passed the exam and am currently working as Jr. Data Scientist at DenizBank.

What Else: Basketball Analytics

I played basketball for ten years and I also trained players. It was always in my mind to do something close to basketball. I saw Formula 1 (F1) analytics account that does F1 analyses, and it inspired me to do something similar with basketball data. You can follow my LinkedIn or basketball analytics account if you’re interested.

Well, that’s pretty much the full story. Thank you a lot for reading it, hope you enjoyed it. If you did, you may consider following or check my other stuff here. If you want to contact me, you can do so via my website.

CLIP, LLaVA, and the Brain

Deep Learning and the Brain Insights into Multimodal Transformers from Neuroscience Jonathan R. Williford, PhD · Follow Published in Towards Data Science · 8 min

Read More »