Priors for Deep Networks: Limit theorems, pitfalls, open questions
1 523
33.8
Microsoft Research333 тыс
Следующее
20.04.18 – 5 81725:19Популярные
91 день – 4071:44
281 день – 2 5543:17Опубликовано 20 апреля 2018, 23:15
Much research in Bayesian Deep Learning is about approximating the posterior. With some notable exceptions, the choice of prior is less often considered. In this second direction, we discuss recent work on central limit theorems for neural networks with more than one hidden layer, some thoughts about over-confident extrapolation, and the dangers of improper priors from the literature.
See more at microsoft.com/en-us/research/v...
See more at microsoft.com/en-us/research/v...
Свежие видео
1 день – 6 1682:30
7 дней – 1 4834:55
10 дней – 7 4391:22
12 дней – 115 02116:43
13 дней – 2 2633:27Случайные видео
50 дней – 6500:48
241 день – 619 9530:41
296 дней – 231 4331:42
24.08.22 – 433 30253:59
07.09.21 – 22 67651:28
03.11.13 – 3 0232:00
1 день – 2 25618:45
7 дней – 5 0361:00
8 дней – 2511:52
15 дней – 11 8320:33
16 дней – 38456:09
16 дней – 1 7283:27
