@Cambridge_Uni Grad - Astrophysics | @join_EF alumni | a tired MedTech founder and entrepreneur | hoping to take a break and do PhD in applying 
to 
Możesz dodawać lokalizację do Twoich Tweetów, jak miasto czy konkretne miejsce, z sieci lub innych aplikacji. W każdej chwili możesz usunąć historię lokalizacji swoich Tweetów. Dowiedz się więcej
We address the question in depth in the paper. It's actually quite subtle and hard to give a short take. They're mostly useful when you have a small number of observations (e.g. BayesOpt with a really expensive objective). But there is more to the story.
In 5.1.1 you discuss the LL and MSE for TP vs GP. The results on spatial and wine data were particularly noteworthy; with the TP having lower MSE scores and ridiculously high LL scores. I wasn't sure why this was the case.
It's a different model. The uncertainty representation is different. After optimizing the marginal likelihood for each, the kernel hypers will also be different. The noise model is subtly different. The paper discusses the differences.
Got it, thanks! That clears things up. I'll spend more time carefully re-reading the discussion to get a better understanding of the subtle differences
Twitter jest przeciążony lub wystąpił chwilowy problem. Spróbuj ponownie lub sprawdź status Twittera, aby uzyskać więcej informacji.
