My educational summer time

posted by Nikolay Baldin on September 28, 2015

I attended a few very exciting events this summer: two conferences and one summer school (all in stats). Below you will find my experience wandering the streets of Nice, eating seafood in Amsterdam and exploring an ideal place for doing research. Among forthcoming events in stats, I would highly recommend to apply to the spring school organized by Markus Reiß and Natalie Neumeyer with more information to appear soon on this website.

Summer school in statistics in Frejus, France. Aug 31 - Sep 4

The courses were given by Emmanuel Candes on multiple testing and selective inference and Ery Arias-Castro on inference in networks. In particular, Emmanuel presented his recent work about the knockoff filter. I kind of admire the way he presents his work.. I really enjoyed the event and learned a lot about multiple testing (probably, will study a few papers). Most of the things Ery presented I more or less had a good picture of and so wasn't an active listener.

The location was great. I also spent a day with Nazar wandering the streets of Nice, Frejus and Saint-Rafael.

There were around 70 participants with 90 % of them being French. The program was rather tight: courses in the mornings starting at 9.15 and PhD talks (around 35) in the evenings till 19.20. I also gave a talk about volume estimation. At this school, I realized that PhD sessions talks are virtually useless for several reasons: 1. No seniors usually attend them 2. Only 30-40 % of students themselves attend other students' talks :) As a result, you have almost no feedback at the end. I must confess that I also missed many PhD talks swimming in the sea..) In general, it seemed like the organizers did not really care about providing students with a platform for introducing and getting to know each other. The excursions do not count. It's quite funny that when you join a group of French students talking with each other they don't necessarily switch to English ..)

Bonus pic :)

To wrap up, I really encourage my dear reader interested in stats to apply to the next spring school organized by Markus Reiß and Natalie Neumeyer. They provide lots of real opportunities for communication including short (!) PhD students' introductions at the beginning, poster sessions and tutorials. This year there are even more great features to come! Although, it has not been announced yet, this year school is taking place near Berlin with 3 top-level specialists in statistics including Alexandre Tsybakov, Ramon van Handel and Harrison Zhou (I may be wrong though). More information should appear on this website. Stay tuned.

European meeting of statisticians in Amsterdam. July 6-10

That's a huge conference in statistics held at Vrije Universiteit Amsterdam. Lots and lots of talks in various areas of statistics. I realized that I don't really understand such big meetings and probably won't attend any in the future. It could easily happen that you meet an interesting person at one session and then you never see him.. Among talks, I enjoyed the session Computational and Statistical Aspects in High-Dimensional Inference with Ankur Moitra, Phillipe Rigollet and Emmanuel Candes. I also gave a 20 minutes invited talk in front of a small group of people (many of them seemed to be interested though).

The food and coffee breaks were bad with no dinners provided. I also didn't feel myself comfortable in the university. The one good thing is that they organized good reception events and granted vouchers to a canal cruise for participants:) In general, I got a bad impression about the conference, which is quite sad especially because I preferred EMS to COLT held in Paris at the same time.

Amsterdam is cozy and cute.

Among restaurants, I liked The Seafood Bar" and "Cannibale Royale" (thanks Suzanne for recommending the last one :)). The food is pricey. It was difficult to find a good coffee in the city center. With Sasha, we also rent bicycles at Starbikes Rental to explore the city. The bicycle traffic is really chaotic and dangerous (missed Berlin). I was buffled that motorcycles freely use the same paths as bicycles..

Oberwolfach meeting. "Probabilistic Techniques in Modern Statistics". May 17-23

That's really a dream place for conducting research.. Absolutely stunning and fecund environment.. It's funny that I got there by chance: one of the registered participants couldn't come and so I took his place (thanks to kind organizers for the opportunity). As a result, I was invited 3 days before the conference started, but my suitcase is always ready..)

There were around 30 talks given by world-class specialists in probability theory and statistics. It was actually the first conference where I attended every single talk. The most memorisable talks for me were by John Duchi, Ramon van Handel, Johannes Schmidt-Hieber, Jean Jacod and Harrison Zhou to mention just a few. The report is available here. There was also a small PhD session consisted of 4 talks including mine. I was impressed that many senior researchers attended the session and I gained a good feedback.

In Oberwolfach, there were many small details that got me really excited about the place. You don't really understand how unique that place is until you come there. First, all rooms including participants' rooms are always open (unless they are locked from inside). Next, the distribution of names at lunches and dinners is random and furthers communication among participants (I think this is what many schools need and should adopt) etc.

Unforgettable experience. Hope to come back in the future..)