Yeah, I'm sure there is literature here, I just need to dig a bit! I've been doing some quite crude thresholding so far to decide when things are low-rank 'enough' for the switch. This works sometimes, but seem to also interfere with adaptation of my HMC sampler

Working on something very much in this vein right now actually. I have a super high-dim problem where due to sparsity the problem becomes low rank during sampling. Wanting to then switch from doing inverses via Cholesky to inverses with Woodbury -- but quite fiddly numerically so far...

oh thanks! my role was tiny though, the first author (a master's student (!)) is fantastic and did the work

At a conference this summer a guy referred to papers from the late 90s as «turn of the last century»…

Been a while since I looked into this but the Hutch++ paper has some optimality results I think as well as some refs to variance reduction techniques arxiv.org/abs/2010.09649 ( from 2021 though, so I'm sure a lot has happened since )

I’d also be interested in this, Martin. How it is structured, what is covered, what the pre-requisites are etc

I think that could indeed be quite illuminating. Often we approach these topics trying to make them practical without requiring too much mathematical setup. You don’t *need* stochastic process theory to get at the core of GP regression, but you might need it to understand the finer details later on

Interesting post! I helped run a similar module in Cambridge last year were we tried to cram in too much, covering basics of GPs *and* DPs. If the emphasis is on BNP I think one must mention DPs at some point, though it is a step up in abstraction compared to GPs. «distribution of distributions» etc

It can be bounded from below and above, by \sqrt{n} and n, respectively. These cases each reflect say an extreme lengthscale, ell=\infty or ell=0, and so it appears almost like a measure of complexity. I'm struggling to come up with an intuitive explanation.

Does anyone have any ideas/references?

Maybe I'll add that the call is very general across "computational science", encompassing biology, chemistry, physics etc. alongside maths & stats. Bayesian machine learning is just one of the projects within our department, see the link below
www.uio.no/dscience/eng...

Are you assuming you are able to compute the derivatives as well, or simply that you know they are non-negative?

Overbygningen er vel laget av stål..?

It's quite ironic that the term was first coined in a sociological essay warning *against* meritocracy, rather than lifting it up as an ideal.