There is no room of strangers that I’d rather be in than that full of curious minds.

Today I found myself a bit over my head. Loving everything statistics I sat in a talk that turned out to be more on the heavy theoretical side, rather than the applied side, that I’m more comfortable with.

I did find the format of the discussion and my conversations with the all-academic participants very interesting.

The talk was hosted by the Royal Society of Statistics in Morgate area in London (fun fact: Reverand Bayes is buried right around the corner). Two authors Sebastian Engelke and Adrien Hitz presented their new paper, afterwhich participants in the crowd followed with their own presentations about the said paper. The participants, each in turn, congratulated the authors on their achievement, explained why it was innovative, timely, applicable as well as pointing out limitations and need for future research. A few of them even hailed it a New subject in statistics. Very amicable indeed! About four or five researchers gave their constructive criticism. This was followed by comments read by one of the chair of eight more who couldn’t fly in or take the train. The authors had the last say where they got to give preliminary response to the said limitations and promissed to follow up with a more thoughtful repsonse to add to all the remarks that will be published in an upcoming journal of the RSS. I learned from the chair of this discussion that this format has been going on for centuries!

I later joked with the authors and the other professional statisticians that what attracted me most was my misreading of the abstract, in which they mentioned Pareto distributions (a type of distribution that decays like the exponential just with a heavier tail) which I confused for Pareto optimisation (a method of optimisating a multiple parameters), both named after an economist from the 19th century. I did have some personal take-aways. The core of the talk was how to use graphical models to quantify for exterme outliers, which is useful, e.g, for weather forcasting.

Like I said, the technical details were a bit above my head, but I did enjoy speaking with the professional statisticians about their work and life in academia. Sharing “war” stories of my days as an academic doing cosmology research as well as a Data Scientist in the private sector. From the statistians I heard the familiar stories, e.g, the the two body problem (partners finding jobs in the same city), but also more interesting ones of how their work is used for government policy, e.g, fishing quota optimisation in Canadian waters.