Could you tell us a little about yourself and about your studies before coming to Imperial?
Before coming to Imperial, I studied mathematics at Oxford University. During my time there, I enjoyed probability and statistics the most, and took several courses in that area, including an introduction to statistical machine learning in my final year. Although my undergraduate degree was very much theory-oriented, with a “pen-and-paper” style of exercices, I had the chance to complete a few internships. These included time at NukkAI, an AI lab, where I started to program and apply some of the machine learning concepts I had discovered in my undergraduate degree.
The thing that excited me the most about artificial intelligence was how it couples strong mathematics and statistical theory with the power of computing. Eager to learn more, I decided to apply to master’s programs.
What attracted you about the MSc in AI?
What attracted me the most with Imperial’s MSc in AI was the school’s strong reputation in STEM and the variety of modules on the degree. Unlike most master's in artificial intelligence, Imperial’s does not focus merely on machine learning, but also includes modules on symbolic AI, a less-popular yet important paradigm of artificial intelligence which tries to reason with symbols and propositions using logic. I also liked the fact that the program included a software engineering project. I participated in a few hackathons prior to joining Imperial and since then, software engineering has always been something I wanted to explore and get better at.
What did you enjoy the most?
Many things come to my mind as I read this question! I remember particularly enjoying implementing some of the reinforcement learning algorithms discussed in lectures, playing around with GANs and VAEs to generate artificial images, and learning about crypto-currencies in the Principles of Distributed Ledgers module. All of the modules were super fun and nicely intertwined theory and practice.
I also enjoyed a lot meeting everyone else in the programme. The cohort was very diverse, with people from all around the world with all kinds of backgrounds, and it created a really nice environment. Everyone was also super helpful and motivated, making the learning experience even more enjoyable.
What did you find more challenging?
What I found most challenging was the number of deliverables we had throughout the year. The AI master's is for sure very intense, but you just learn so much that it is definitely worth it!
Could you tell us about some of your achievements on the MSc that make you proud?
The Web3 Application that my team and I built as part of the software engineering project is what makes me most proud. This was definitely very challenging, since many of us were new to the languages and technologies we had to use. For example, we had to learn how to write and deploy Ethereum smart contracts, code our front end in Javascript using React JS, and connect our application to IPFS, a decentralised storage system. This was very rewarding though—delivering, by the end of the module, a functional end-to-end web application.
What did you do in your spare time?
In my spare time, I enjoyed meeting up with friends, playing football and going out bouldering. Although the workload was high, you do get a chance to rest and have fun!
Could you tell us about your individual project?
In my individual project, I focused on developing probabilistic machine learning methods to take into account uncertainty in image registration, an image processing technique used for example in image segmentation, an application I explored in the later part of my thesis.
I think the project is a great opportunity to further explore a topic of your choice. There is a huge list of projects you can choose from, and you can even come up with your own if you feel like it! In my project, I got to propose, implement and test my own ideas which was very exciting. My supervisor was also really helpful, providing guidance all along the way.
What have you been doing since you graduated?
A couple of months after graduating, I joined Google as a Software Engineer. My team works on AdSense, a product that helps website owners monetize their content. Although I don’t do ML on a regular basis, my work is super fun and full of challenges.
Do you have any advice for prospective students?
My advice is to be open to try new things. There are so many great modules that you can do in this master's, not all of them related to Deep Learning! Personally, I took Principles of Distributed Ledgers and it ended up being my favourite.