Winning teams back from China to share their experiences
Miquel Larsson Corominas and Elena Florentina Ionescu are back from their one week trip to China. They belonged to the winning teams of a Hackathon hold between the EECS-school and Technology company Huawei. The goal of the Hackathon was to process more than 100k unlabeled data and propose an algorithm able to generalize and compress unlabeled information.
Why did you choose to study at KTH?
Elena: "KTH is one of the most prestigious universities in Europe. Its high ranking and great feedback from the students made me chose this university."
Miquel: "Studying at KTH was a long-time dream of mine, as one of the top technical universities in Scandinavia. Then, the master in machine learning was a perfect fit for what I was looking for. It turns out that completely dedicated masters in machine learning are a rare exception."
Why did you attend this hackathon?
Elena: "I am passionate about AI and Deep Learning, but I don’t have many courses in my curricula with these subjects, so I got excited when I saw the hackathon. Also, I’ve never been before outside of Europe, so the trip to China draw my attention!"
Miquel: "I love going to hackathons, and always try to go to as many as I can, as they are a great opportunity to push your limits and learn a lot. I highly recommend to anyone that has motivation to learn to attend these events, they are great. In the worst case scenario, you will meet a bunch of people with your same interests and will realize what knowledge you are missing."
What use did you have of your skills in the hackathon?
Elena: "I had experience with participating in Machine Learning hackathons, therefore handling data and analysing it was very useful in order to figure out what algorithms we can use. Moreover, from a theoretical point of view I was familiar with Autoencoders (our first approach) and I used algorithms fore creating embeddings and classifications both in my ML courses and for my bachelor thesis project."
Miquel: "The hackathon was related to deep learning, which was something I had studied on a theoretical level, but had never had to actually use on a practical level as required in the competition."
How difficult was the challenge which you were given?
Elena: "I think it is hard to create a good solution is a short amount of time especially when it is required to use certain technologies. I believe that the challenge was hard, but approachable."
Miquel: "The challenge had a perfect difficulty level, in my opinion. The technical aspect of it was adequate, and the limited timeframe we had (24 hours) made it much more interesting, as the organizing capacity of the team would be critical in order to achieve some result."
What did you appreciate the most from the mentors?
Elena: "We received a lot of support not only from a technical point of view but also they kept our moral up through the night and in the morning before the presentation."
Miquel: "With such a limited time for such a complex task, mentors being available to quickly solve any questions that arose with the data and the platform that was provided was great and super helpful."
What was the best and most most exciting part of the hackathon?
Elena: "Seeing the production line with automated robots. I worked on projects with computer vision and machine learning that was building software for these robots, but I’ve never seen an automated production line before! Also, seeing the second tallest building in the world was stunning! And when we submitted our result in the leader board and we saw that we are in top 3."
Miquel: "Without any doubt, working along my team in order to give our best and of course, when your team is announced as a winner is one of the most exciting moments. But also, the moment of submitting a solution the team is proud of is super rewarding."
What did you learn during your trip to China?
Elena: "From a cultural perspective, every moment of the trip was a learning point for me. Also, I was not aware of the broadness and high quality of all the services that Huawei provides. I had the feeling that everything that can be done in the communications field, it is done at Huawei. It was very insightful to talk with doctorates that were presenting their work within the company and the cutting-edge research that is done here."
Miquel: "It was a once in a lifetime opportunity, which I am extremely happy to have been able to enjoy. Being able to see the offices and factories of Huawei in person was amazing, which combined with visiting the surrounding cities and tasting all the different gastronomy made it an incredible experience."
What will you take with you from this experience?
Elena: "That China lives in the future, and not only with the 7 hours time-zone difference, but with all the technology that is built there."
Miquel: "A lot more about how to work with neural nets in a practical way. And of course, how to manage a team a bit better!"
What are you doing next at KTH?
Elena: "I am going to finish my studies at KTH in the beginning of summer 2019."
Miquel: "This spring will be my final semester, in which I will be working on my master thesis. It will actually be related very closely to what we did in the hackathon: unsupervised feature extraction."