# Top tagging with deep neural networks \[Vidyo\]

- Institution: CERN
- Department: Machine Learning
- Speaker: Jannicke Pearkes
- Year: 2017
- Language: English
- Subjects: Computer science
- Source: [CERN](https://videos.cern.ch/record/3017305)
- 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)

Recent literature on deep neural networks for top tagging has focussed on image based techniques or multivariate approaches using high level jet substructure variables. Here, we take a sequential approach to this task by using anordered sequence of energy deposits as training inputs. Unlike previous approaches, this strategy does not result in a loss of information during pixelization or the calculation of high level features. We also propose new preprocessing methods that do not alter key physical quantities such as jet mass. We compare the performance of this approach to standard tagging techniques and present results evaluating the robustness of the neural network to pileup..

0:00:00 Slide 1

0:00:37 Slide 2

0:00:59 Slide 3

0:01:20 Slide 4

0:03:03 Slide 5

0:04:13 Slide 6

0:05:31 Slide 7

0:06:24 Slide 8

0:06:30 Slide 9

0:07:18 Slide 10

0:07:46 Slide 11

0:08:15 Slide 12

0:08:29 Slide 13

0:08:54 Slide 14

0:09:08 Slide 15

0:09:54 Slide 16

0:10:14 Slide 17

0:10:48 Slide 18

0:11:01 Slide 19

0:11:21 Slide 20

0:12:30 Slide 21

0:13:29 Slide 22

0:13:35 Slide 23

0:14:22 Slide 24

0:14:41 Slide 25

0:14:56 Slide 26

0:16:06 Slide 27

0:16:30 Slide 28

0:17:54 Slide 29

## Lectures

1. [Top tagging with deep neural networks \[Vidyo\]](https://invisiblenetwork.ai/learn/cern/3017305/e1063321229a), 23 min (CERN's player)