from Jakob Bernoulli to Hong Kong

Here are my slides (or at least the current version thereof) for my talk in Hong Kong at the 2013 (59th ISI) World Statistical Congress(I stopped embedding my slideshare links in the posts as they freeze my broswer. I wonder if anyone else experiences the same behaviour.)

This talk will feature in the History I: Jacob Bernoulli’s “Ars Conjectandi” and the emergence of probability invited paper session organised by Adam Jakubowski. While my own research connection with Bernoulli is at most tenuous, besides using the Law of Large Numbers and Bernoulli rv’s…,  I [of course!] borrowed from earlier slides on our vanilla Rao-Blackwellisation paper (if only  because of the Bernoulli factory connection!) and ask Mark Girolami for his Warwick slides on the Russian roulette (another Bernoulli factory connection!), before recycling my Budapest slides on ABC. The other talks in the session are by Edith Dudley Sylla on Ars Conjectandi and by Krzys Burdzy on his book The Search for Certainty. Book that I critically reviewed in Bayesian Analysis. This will be the first time I meet Krzys in person and I am looking forward to the opportunity!

8 Responses to “from Jakob Bernoulli to Hong Kong”

  1. Dan Simpson Says:

    Nice slides! It’s an interesting route through simulation-based inference.

    On the MCMC and beyond slide, VB should probably be there with EP and INLA. It might also be cool to mention some stuff of MCMC that targets the wrong distribution (it may be somewhere in the acronym soup – sorry! I can never keep these things straight!)

    • Dan Simpson Says:

      Do you know if the inexact MCMC methods suffer a curse of dimensionality? It doesn’t “feel” like they do beyond that of the randomized MCMC chain that its coupled to…

    • Dan Simpson Says:

      Apology for the endless stream of comments. The link between R-B and the Bernoulli factory is a bit obscure for me. What is f?

      • Dan Simpson Says:

        And again! Should it be a pi_epsilon at the bottom of the noisy MCMC slide, or does It actually make an exact algorithm?

      • yes and no, in that this is the true posterior distribution for the noisy model. so yes it depends on epsilon and no it is not an approximation…

      • Thanks, Dan! I was certainly a wee bit hasty: f in the Bernoulli factory problem is an arbitrary known function with values in [0,1]. Omiros has an example in his slides about a Brownian bridge…

  2. Yes I also have problems (with firefox) when having a tab containing a slideshare presentation open: it creates a gigantic memory leak. If I remember correctly Google Chrome is not affected (from the top of my head, but I am not sure…)

Leave a comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.