Trading Convexity for Scalability
151
Microsoft Research330 тыс
Следующее
Популярные
251 день – 6784:47
335 дней – 1 3611:52Опубликовано 6 сентября 2016, 6:32
Joint work with Ronan Collobert, Fabian Sinz, and Jason Weston. Convex learning algorithms, such as Support Vector Machines (SVMs), are often seen as highly desirable because they offer strong practical properties and are amenable to theoretical analysis.  However, in this work we show how non-convexity can provide scalability advantages over convexity.  We show how concave-convex programming can be applied to produce (1)~faster SVMs where training errors are no longer support vectors, and (2)~much faster Transductive SVMs.
Свежие видео
5 дней – 52 08013:35
7 дней – 96 50719:30
13 дней – 2790:47
14 дней – 5880:11
17 дней – 138 65811:17Случайные видео
108 дней – 21 4080:20
167 дней – 2 10326:53
172 дня – 37 1550:31
322 дня – 3 1180:42
03.07.22 – 126 7895:44
06.11.08 – 36 35612:22
6 дней – 4433:31
7 дней – 2 801 2381:34
8 дней – 2351:57
8 дней – 40 83623:30
10 дней – 2 6045:10
13 дней – 202 2376:39
14 дней – 56 59312:02
