Next Article in Journal
An Emotional Model Based on Fuzzy Logic and Social Psychology for a Personal Assistant Robot
Previous Article in Journal
Time Domain Investigation of Hybrid Intelligent Controllers Fed Five-Phase PMBLDC Motor Drive
Previous Article in Special Issue
Machine-Learning Application for a Likelihood Ratio Estimation Problem at LHC
 
 
Review
Peer-Review Record

VBF Event Classification with Recurrent Neural Networks at ATLAS’s LHC Experiment

Appl. Sci. 2023, 13(5), 3282; https://doi.org/10.3390/app13053282
by Silvia Auricchio 1,2,*,†, Francesco Cirotto 1,2,*,† and Antonio Giannini 3,*,†
Reviewer 1: Anonymous
Reviewer 2:
Appl. Sci. 2023, 13(5), 3282; https://doi.org/10.3390/app13053282
Submission received: 11 February 2023 / Revised: 24 February 2023 / Accepted: 24 February 2023 / Published: 4 March 2023
(This article belongs to the Special Issue Machine Learning Applications in Atlas and CMS Experiments at LHC)

Round 1

Reviewer 1 Report

In this work, A ML approach based on Recurrent Neural Network (RNN) for event topology identification in High Energy Physics (HEP) is presented. The Vector Boson Fusion production mechanism in proton collisions is predicted both from the current theoretical model of the Particle Physics, the Standard Model, both from new extensions of this model. This process has a well defined event topology and detector signature. In this work, a ML approach based on the RNN dealing with hadronic only event information is developed to enhance the acceptance of this production mechanism. This approach, then, has been applied to a physics analysis in the context of high mass diboson resonances searches in the ATLAS experiment. 

The work is interesting and merits publications in applied sciences after addressing the following few minor comments:

1. Will the number of possible pairs can be more than 15?

2. Elaborate more on the motivation of the work in the introductory section.

3. Include all figures in eps format.

4. Include future extensions in this direction in concluding remarks.

5. Proofread the whole manuscript for typos and grammatical errors.

Author Response

Dear Reviewer,

 

thanks for your comments. You can find in attachment a file containing our replies and the changes in the paper with respect to previous draft.

 

Best regards,

Francesco Cirotto (on behalf of all authors)

Author Response File: Author Response.pdf

Reviewer 2 Report

In fact, the effective classification of final-state signatures in two dominant channels of pp - reactions at the LHC with the help of Machine Learning methods is discussed in this paper. As an important element, it is proposed to introduce Recurrent Neural Network (RNN) into the general architecture of the program separating  specific and substantial events. This approach was tested for the analysis of the two-boson production of semilepton final states at the ATLAS collaboration, it is assumed that  the scheme can be extended to other types of reactions and final states.

The article is of interest because the described methods are currently attracting more and more attention for the analysis of large arrayss of data not only at LHC collaborations, but also at other measuring complexes (LHAASO, IceCube, cosmic telescopes gathering a bulk of astrophysical  data and so on). The algorithm for the RNN using in the general analytical chain is described in sufficient detail.

The minor disadvantages could be attributed to the lack of a more detailed discussion of the applicability and, possibly, some necessary modification of this technique to classify and separate events with other signatures, possibly for the analysis of extended air showers, multi-muon events and others.

In principle, the article can be published in the Applied Science journal.

Author Response

Dear Reviewer,

 

thanks for your comments. You can find in attachment a file containing our replies and the changes in the paper with respect to previous draft.

 

Best regards,

Francesco Cirotto (on behalf of all authors)

Author Response File: Author Response.pdf

Back to TopTop