# Reports of My Demise Are Greatly Exaggerated: $N$-subjettiness Taggers Take On Jet Images

Abstract : We compare the performance of a convolutional neural network (CNN) trained on jet images with dense neural networks (DNNs) trained on $n$-subjettiness variables to study the distinguishing power of these two separate techniques applied to boosted hadronic $Z$ boson and top quark decays. We find that they perform almost identically once jet mass information is included in a consistent manner, which suggests they are accessing the same underlying information which can be intuitively understood as being contained in 4-, 5-, and 6-body kinematic phase spaces depending on the sample. This suggests both of these methods are highly useful for heavy object tagging and provides a tentative answer to the question of what the image network is actually learning.
Keywords :
Type de document :
Pré-publication, Document de travail
2018
Domaine :
Liste complète des métadonnées

https://hal.archives-ouvertes.fr/hal-01851157
Contributeur : Inspire Hep <>
Soumis le : dimanche 29 juillet 2018 - 14:04:58
Dernière modification le : mardi 9 octobre 2018 - 09:07:03

### Citation

Liam Moore, Karl Nordström, Sreedevi Varma, Malcolm Fairbairn. Reports of My Demise Are Greatly Exaggerated: $N$-subjettiness Taggers Take On Jet Images. 2018. 〈hal-01851157〉

### Métriques

Consultations de la notice