If there’s one AI summer school to look out for in 2020, it will be the Probabilistic Artificial Intelligence (ProbAI) school. There’s also an opportunity to win a generous scholarship to participate. This year 2019 was the maiden edition of the AI summer school and held at the Norwegian University of Science and Technology (NTNU) from 4th to 7th June. As new as it was, it received over 400 applications out of which 150 applicants were selected from 43 nationalities. Here are 5 reasons you should start warming up for the 2020 school.
1. The statistics feel
I must warn you that this AI summer school is not targeted at a beginner in statistics. Their website states that an applicant should be intermediary PhD student in computational, mathematical and statistical knowledge for Machine Learning and Artificial Intelligence. Also included are exceptional MSc students and experienced researchers. Bayesian was the buzz word at this AI school. ProbAI would throw a lot of statistics darts that all the phobia for statistics would be gone before it ends. I thought I would be the only one feeling lost in the classes but from my interactions, I realised I had company. My advice for a beginner would be: try to learn the basics before the school and make sure you understand the intuition behind the methods before you finish the course. You could learn more about the need for Bayesian statistics in real-world applications and deep learning. One nice thing was that the core theories were re-emphasized by several of the tutors. This way, I was able to lay hold of the foundation and main ideas. There was a hands-on session and it was a good time to play with the codes and ask questions. There were people in the hall to assist although not too many to serve the entire class.
2. Scholarship award
Since I did not have the luxury of being sponsored for multiple summer schools by my institution, I banked on the possibility of securing a scholarship. It was on this premise that I applied for the course. As most grants, the scholarship was competitive. It was offered to around 12 persons, with a total worth of about 110,000 NOK. I didn’t even make the first shortlist. I was among the 13 applicants selected to be on the waiting list. The scholarship was very generous. You either chose to be paid a reimbursement of eligible expenses after the school or the organisers utilised a portion to make travel arrangements. My award came with a registration waiver and a grant covering transportation, accommodation and dining. The scholarship amount varied depending on which geographical location you came from. This means that those staying closer to Norway will receive less money. The application process was quite simple. I was asked the reason I needed the scholarship and my budget for expenses I wanted covered by the award. It was mentioned on the website that the scholarship focused mainly on those from developing countries and under-represented groups.
3. Coursework (project)
The AI summer school is offered as a PhD course at NTNU with 7.5 ECTS credit points. If you need to do a course as part of your research training, then you could consider doing the project assignment. There’s another reason I would recommend you do the coursework. For me, it was a way to cement what I learnt at the week-long AI summer school. It was at the point I started working on the coursework that I began to appreciate better how useful the course would be, not only for my research but also for my future career. It has also helped me go back to the drawing board to understand certain concepts I had only been regarding as a black-box. The likes of backpropagation, optimisation, gradient descent. Sometimes one just briskly learns things to keep going when they haven’t yet settled in one’s head.
4.Trondheim twilight
I was really ‘WOWed’ at Trondheim. The natural scenery left me in so much admiration for the city. On the wednesday we went hiking as a group. We commenced hiking around 8pm and I wasn’t home until 1am. I came to see first hand what ‘it doesn’t get dark here during summer’ really meant. There wasn’t just enough time to explore the city as the daily classes concluded when the main attractions like the art museums and cathedral had already closed. I managed to steal a few minutes during one of the break to visit the Nidaros cathedral. Another fascinating experience was a tour to the University’s computer science museum.
5. Because it can only get better
The program was very well organized especially for a first time event. Kudos to the organisers! The daily lunch was an exquisite buffet, and we had different variety of meals each day. The evening banquet and welcome reception were great initiatives for a social gathering. I was inspired to see fellow PhD student like Thomas Lucas handle the lecture session. Personally, this motivates me to learn more knowing a colleague can do this. One of the feedback I gave about the event was for the organizers to give thought to having a dedicated time and space for the poster sessions in the program. I feel this is better than having the posters behind the class where they almost go unnoticed. I presented my poster on the Wednesday of the week. Since there were no slots for the posters, it was only during coffee breaks that people got the chance to visit the posters. This was too brief to have a meaningful interaction. I so much believe it can only get better for the 2020 event as they pulled off an amazing one for a first.
Finally
Norway is a very student and research friendly environment. It even reflects from the point of applying for visa. If you’re an international student or researcher, Norway doesn’t charge you visa fees. You would pay the fee during the application but they will refund you. The country is very expensive to live in though; so don’t make plans of shopping there.
If you have any questions regarding the course and my experiences during the summer school I will be glad to answer.