| | |
| | |
Stat |
Members: 3645 Articles: 2'504'928 Articles rated: 2609
25 April 2024 |
|
| | | |
|
Article overview
| |
|
A jet tagging algorithm of graph network with HaarPooling message passing | Fei Ma
; Feiyi Liu
; Wei Li
; | Date: |
25 Oct 2022 | Abstract: | Recently methods of graph neural networks (GNNs) have been applied to solving
the problems in high energy physics (HEP) and have shown its great potential
for quark-gluon tagging. In this paper, we introduce an approach of GNNs
combined with a HaarPooling operation, called HaarPooling Message Passing
neural network (HMPNet). The information of jet events is converted into graph
representation as input of HMPNet and the output discrimination scores give the
results of quark-gluon classifications. In HMPNet, the Haar matrix passes
additional information on particles in the process of message passing neutral
network (MPNN), so that the features contain more raw information with updating
during training. This information is embedded into the Haar matrix by Haar
basis, obtained by clustering of $k$-means sorting by different particle
observables. We construct the Haar basis from three different observables:
absolute energy $log E$, transverse momentum $log p_{T}$, and relative
coordinates $(Deltaeta,Deltaphi)$, then discuss their impacts on the
tagging and compare the results with using MPNN and ParticleNet (PN). | Source: | arXiv, 2210.13869 | Services: | Forum | Review | PDF | Favorites |
|
|
No review found.
Did you like this article?
Note: answers to reviews or questions about the article must be posted in the forum section.
Authors are not allowed to review their own article. They can use the forum section.
browser Mozilla/5.0 AppleWebKit/537.36 (KHTML, like Gecko; compatible; ClaudeBot/1.0; +claudebot@anthropic.com)
|
| |
|
|
|
| News, job offers and information for researchers and scientists:
| |