# Less is more: regularization perspectives on large scale machine learning

- Institution: CERN
- Department: EP-IT Data science seminars
- Speaker: Lorenzo Rosasco
- Year: 2017
- Language: English
- Subjects: Computer science
- Source: [CERN](https://videos.cern.ch/record/3004614)
- Licence: [CERN](http://copyright.web.cern.ch)
- Who can watch: Tweens profiles and older (sources give no ratings; placed by Invisible Network from the source's own words)

Modern data-sets are often huge, possibly high-dimensional, and require complex non-linear parameterization to be modeled accurately.

Examples include image and audio classification but also data analysis problems in natural sciences, e..g high energy physics or biology.

Deep learning based techniques provide a possible solution at the expanse of theoretical guidance and, especially, of computational requirements. It is then a key challenge for large scale machine learning to devise approaches guaranteed to be accurate and yet computationally efficient. In this talk, we will consider a regularization perspectives on machine learning appealing to classical ideas in linear algebra and inverse problems to scale-up dramatically nonparametric methods such as kernel methods, often dismissed because of prohibitive costs. Our analysis derives optimal theoretical guarantees while providing experimental results at par or out-performing state of the art approaches. .

0:00:00 Slide 1

0:02:17 Slide 2

0:02:54 Slide 3

0:03:34 Slide 4

0:04:26 Slide 5

0:04:56 Slide 6

0:05:56 Slide 7

0:06:59 Slide 8

0:08:08 Slide 9

0:09:56 Slide 10

0:10:11 Slide 11

0:10:48 Slide 12

0:11:01 Slide 13

0:11:39 Slide 14

0:11:51 Slide 15

0:12:57 Slide 16

0:13:59 Slide 17

0:14:04 Slide 18

0:14:32 Slide 19

0:16:52 Slide 20

0:16:56 Slide 21

0:17:00 Slide 22

0:17:20 Slide 23

0:17:46 Slide 24

0:19:00 Slide 25

0:19:51 Slide 26

0:20:31 Slide 27

0:23:43 Slide 28

0:24:36 Slide 29

0:25:46 Slide 30

0:25:54 Slide 31

0:26:03 Slide 32

0:28:49 Slide 33

0:29:29 Slide 34

0:29:33 Slide 35

0:29:50 Slide 36

0:30:39 Slide 37

0:31:52 Slide 38

0:32:26 Slide 39

0:33:00 Slide 40

0:33:34 Slide 41

0:33:43 Slide 42

0:34:07 Slide 43

0:34:57 Slide 44

0:35:22 Slide 45

0:37:11 Slide 46

0:37:30 Slide 47

0:38:23 Slide 48

0:39:03 Slide 49

0:39:43 Slide 50

0:40:05 Slide 51

0:40:53 Slide 52

0:41:22 Slide 53

0:41:29 Slide 54

0:42:07 Slide 55

0:43:02 Slide 56

0:43:49 Slide 57

0:43:59…

## Lectures

1. [Less is more: regularization perspectives on large scale machine learning](https://invisiblenetwork.ai/learn/cern/3004614/063271d9e229), 1 h 6 min (CERN's player)