Students interested in attending and participating in the upcoming ISIPTA meeting (July 2023 in Spain) can apply for financial support to help cover travel expenses. More information is available here.

There are also prizes awarded at each ISIPTA meeting for best student papers, which is a nice honor (and a nice monetary prize). My former student, Leonardo Cella, was one of the winners at ISIPTA 2021, with his paper Valid Inferential Models for Prediction in Supervised Learning Problems. Leo was also invited to submit an extended version of the conference paper to IJAR that was eventually published (preprint and journal version). My point is just that this kind of recognition isn’t impossible — unless you don’t participate.

Christian Frohlich at the University of Tubingen, Germany, shared with me his paper with the above title. I’ve only skimmed it so far but this seems like an interesting alternative to the way I’ve been trying to infuse imprecise probabilities with statistical/frequentist properties.

Students, I hope you’re starting to make plans for the course project. Again, this is basically a literature review on a topic of interest that relates to imprecise probability in some way. A working list of potential topics for the project is given on the Project page on the course website.

Below is the project timeline, including the two checkpoints prior to the due date where I’d be able to offer some feedback, suggestions, etc.

1. A project proposal is due to me via email on Friday, October 14th. This is just a short document that lists the general topic that you plan to investigate, a brief explanation of why you’re interested in or why you chose this topic (e.g., the topic’s connection to the stat/ML research project you are or hope to be working on for your PhD), and a list of a few (3–5) references that you’ve found to get your literature review started. If you’re working in pairs, then only one proposal document (with both group members’ names) needs to be submitted.
2. By Friday, November 4th, students should send me an email with an informal update on their progress. This is just to make sure that everything is going smoothly, that you’ve found references, and you have a plan of what to include in your project write-up. If you’re stuck in some way, then I can try to help.
3. The project write-up is due to me via email by noon on Monday, December 12th.

The third homework set has been posted, see the homework page for details.

SIPTA recently started hosting semi-regular online seminars on imprecise probability-related topics. The website is here, which lists past talks (and links to the videos on YouTube) and information about how to join for future talks.

Note that 15:00 CEST is 9:00am EST.

The next seminar (Thursday, September 29th) is given by Peter Grunwald on the same topic he presented in the NC State Stat Department Seminar just last week. It was an interesting talk, so it doesn’t hurt to watch a rerun if you’ve already seen it, but I imagine that the details will be tailored to the imprecise probability community who’s hosting the event. There could also be some interesting course project ideas that come out of Peter’s talk…