Bruno Gavranović testimonial

My name is Bruno Gavranović, I’m a Computer Science student from Croatia and I’d like to describe to you how I went from not being accepted to PSIML to actually becoming a lecturer there! Career wise – applying for the Summer Institute turned out to be one of the best decisions I’ve made – it jump-started me into neural networks.

I first heard about Petnica when I was just getting started with Machine Learning. I applied to PSIML back in 2015 – it was the first edition of Summer Institute. I knew how to code, and I’ve heard about Machine Learning - but nothing more. I even messed up the homework assignment and didn’t get in! Some of my friends did, though, and it made me really jealous.

Next year came around. I had more skills and I understood that Machine Learning is not just a whim, but something that really interests me. I applied again and this time – I managed to get in!

PSIML was everything I expected it to be, and more: from the relaxed atmosphere to interesting lectures. We didn’t just listen to lectures – we spent time with lecturers, joked and even had a barbecue! The concept of the seminar itself, where we all spend 10 days together doing something we enjoy, without any distractions is where I think Petnica really shines!

Halfway through the seminar we were assigned team projects. One of the fellow participants and I worked on generating images from scratch with the so-called Generative Adversarial Networks (GANs). It was (and still is) a hot topic – and it was invented just few months before that!

We actually managed to create something that looks like real images. It was such a big morale boost! Turns out we had something that not only works, but also looks amazing!

After PSIML ended, I continued to work on GANs. I had gotten such a good foundation that I was able to work on many different projects in deep learning as well. One of the most notable ones is that I worked on reimplementing DeepMind’s Differentiable Neural Computer, which is regarded as a really sophisticated and complex neural network architecture. Working on university project related to Machine Learning was also easier – I’ve heard about most of the things already.

By that time, I was applying to the AI Grant - a semi-annual project where grants are given to promising AI researchers. I managed to become the AI Grant finalist – which means I got into the top 10% of the applications!

As another PSIML was announced, I applied again! I wasn’t considering that there isn’t more to learn, as a matter of fact – the field grew even larger and I felt like I needed to catch up! But turns out –I had obviously shown a good knowledge of ML that, to my surprise, Đorđe, Momčilo and Goran invited me to give a talk on this year’s PSIML!

I was surprised – “Me?” – I thought.

And that seemed to be it! From not being sure will I get in, I found myself in the situation where I was actually a lecturer. I had worked really hard to try to summarize my knowlage and present this fascinating idea. It turned out amazing – even better than I thought! The Summer Institute was again full of very bright people and I still felt a little bit like a participant!

In summary, I think PSIML is an excellent opportunity for anybody interested in ML. Not only you’ll learn many exciting new things, you’ll also meet many new interesting people who all share your passion.