Rather than a single point mass, we are now using multiple point masses in good locations, which will be a much better approximation of the BMA integral we are trying to solve. 8/18
-
-
Is p(w|D) even a meaningful term when w doesn't refer to parameters of the generating process for D?
- Još 6 drugih odgovora
Novi razgovor -
-
-
This thread is *so* good Andrew. Thanks for taking the time to spell everything out. Happy New Year!
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi
-
-
-
Where are the demos and results of this tangential foray in Bayesian Deep Learning? I like proofs about optimality by contradiction, cases, or induction or even implementation. Is there a palpable demonstration of why I should follow this school of thought rather than the norm?
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi
-
-
-
@threadreaderapp Please compile -
Happy Holidays! Ye yuletide unroll: Thread by
@andrewgwils: Bayesian methods are *especially* compelling for deep neural networks. The key distinguishing… https://threadreaderapp.com/thread/1210354001041969152.html …. See you soon.
Kraj razgovora
Novi razgovor -
-
-
Hi, you can read it here: Thread by
@andrewgwils: Bayesian methods are *especially* compelling for deep neural networks. The key distinguishing… https://threadreaderapp.com/thread/1210354001041969152.html …. Enjoy :)
Kraj razgovora
Novi razgovor
Čini se da učitavanje traje već neko vrijeme.
Twitter je možda preopterećen ili ima kratkotrajnih poteškoća u radu. Pokušajte ponovno ili potražite dodatne informacije u odjeljku Status Twittera.