Deep neural networks for energy and position reconstruction in EXO-200

Delaquis, S.; Jewell, M. J.; Ostrovskiy, I.; Weber, M.; Ziegler, T.; Dalmasson, J.; Kaufman, L. J.; Richards, T H; Albert, J.; Anton, G.; Badhrees, I.; Barbeau, P. S.; Bayerlein, Reimund; Beck, Diane M.; Belov, V.; Breidenbach, M.; Brunner, T.; Cao, G. F.; Cen, W. R.; Chambers, Claire; Cleveland, B. T.; Coon, Minor J.; Craycraft, A.; Cree, W.; Daniels, T.; Danilov, M.; Daugherty, S. J.; Daughhetee, J.; Davis, J.; Mesrobian-Kabakian, A. Der; DeVoe, R.; Dilling, J.; Dolgolenko, A.; Dolinski, M. J.; Fairbank, W. M.; Farine, J.; Feyzbakhsh, S.; Fierlinger, P.; Fudenberg, D.; Gornea, R.; Gratta, G.; Hall, C.; Hansen, E. V.; Harris, D. A.; Hoessl, J.; Hufschmidt, P.; Hughes, M.; Iverson, A.; Jamil, A.; Johnson, Andrew; Karelin, A.; Koffas, T.; Kravitz, S.; Krücken, R.; Kuchenkov, A.; Kumar, Krishna; Lan, Y.; Leonard, D. S.; Li, G. S.; Li, S.; Licciardi, C.; Lin, Yi-Hsuan; MacLellan, R.; Michel, Thilo; Mong, B.; Moore, David C.; Murray, K.; Njoya, O.; Odian, A.; Piepke, A.; Pocar, A.; Retière, F.; Robinson, A.; Rowson, P. C.; Schmidt, Susann; Schubert, A.; Sinclair, D.; Soma, A. K.; Stekhanov, V.; Tarka, M.; Todd, J.; Tolba, T.; Veeraraghavan, V.; Vuilleumier, J. L.; Wagenpfeil, M.; Waite, A. P.; Watkins, J.; Wen, L.; Wichoski, U.; Wrede, Gerrit; Xia, Qi; Yang, L.; Yen, Y.-R.; Zeldovich, O.

doi:10.1088/1748-0221/13/08/p08023

Cited by 56 publications

(49 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Motivated by the results in [26], this analysis introduces a new discriminator for SS and MS events using a DNN that relies on the waveforms of U-wire signals and is found to outperform the searches in [8,20]. The training inputs for the DNN are gray scale images built by arranging neighboring channels next to each other and encoding the amplitudes of U-wire waveforms as pixel values.…”

mentioning

confidence: 99%

Search for Neutrinoless Double-βDecay with the Complete EXO-200 Dataset

Anton¹,

Badhrees²,

Barbeau³

et al. 2019

Phys. Rev. Lett.

Self Cite

212

132

View full text Add to dashboard Cite

A search for neutrinoless double-β decay (0νββ) in 136 Xe is performed with the full EXO-200 dataset using a deep neural network to discriminate between 0νββ and background events. Relative to previous Published by the American Physical Society under the terms of the Creative Commons Attribution 4.0 International license. Further distribution of this work must maintain attribution to the author(s) and the published article's title, journal citation, and DOI. Funded by SCOAP 3 .

show abstract

mentioning

confidence: 99%

Search for Neutrinoless Double-βDecay with the Complete EXO-200 Dataset

Anton¹,

Badhrees²,

Barbeau³

et al. 2019

Phys. Rev. Lett.

Self Cite

212

132

View full text Add to dashboard Cite

show abstract

“…More recently, MicroBooNE considered the problem of classifying experimental data from models trained on simulated data [3]. In addition, CNNs have been shown to be successful in extracting energy and position of events in liquid xenon TPCs [4]. Building on this work, we present results which fine tune pre-trained models, therefore decreasing training time significantly.…”

Section: Challenges Of Data Analysis In the Active-target Time Projecmentioning

confidence: 95%

Machine learning methods for track classification in the AT-TPC

Kuchera

Ramanujan

Taylor

et al. 2019

Nuclear Instruments and Methods in Physics Research Section A:

View full text Add to dashboard Cite

We evaluate machine learning methods for event classification in the Active-Target Time Projection Chamber detector at the National Superconducting Cyclotron Laboratory (NSCL) at Michigan State University. Currently, events of interest are selected via cuts in the track fitting stage of the analysis workflow. An explicit classification step to single out the desired reaction product would result in more accurate physics results as well as a faster analysis process. We tested binary and multi-class classification methods on data produced by the 46 Ar(p,p) experiment run at the NSCL in September 2015. We found that fine-tuning a pre-trained convolutional neural network produced the most successful classifier of proton scattering events in the experimental data, when trained on both experimental and simulated data. We present results from this investigation and conclude with recommendations for event classification in future experiments.

show abstract

“…A key element in the development of machine learning methods is the exploitation of the underlying structure of the data through appropriate architectures. For example, convolutional neural networks make use of local, translation-invariant correlations in image-like data and compile them into characteristic features [1][2][3][4][5][6][7][8][9][10][11][12].…”

Section: Introductionmentioning

confidence: 99%

Lorentz Boost Networks: autonomous physics-inspired feature engineering

et al. 2019

View full text Add to dashboard Cite

A: We present a two-stage neural network architecture that enables a fully autonomous and comprehensive characterization of collision events by exclusively exploiting the four momenta of final-state particles. We refer to the first stage of the architecture as Lorentz Boost Network (LBN). The LBN allows the creation of particle combinations representing rest frames. The LBN also enables the formation of further composite particles, which are then transformed into said rest frames by Lorentz transformation. The properties of the composite, transformed particles are compiled in the form of characteristic variables that serve as input for a subsequent network. This second network has to be configured for a specific analysis task such as the separation of signal and background events. Using the example of the classification of ttH and tt+bb events, we compare the separation power of the LBN approach with that of domain-unspecific deep neural networks (DNN). We observe leading performance with the LBN, even though we provide the DNNs with extensive additional input information beyond the particle four momenta. Furthermore, we demonstrate that the LBN forms physically meaningful particle combinations and autonomously generates suitable characteristic variables.

show abstract

Deep neural networks for energy and position reconstruction in EXO-200

Cited by 56 publications

References 21 publications

Search for Neutrinoless Double-βDecay with the Complete EXO-200 Dataset

Search for Neutrinoless Double-βDecay with the Complete EXO-200 Dataset

Machine learning methods for track classification in the AT-TPC

Lorentz Boost Networks: autonomous physics-inspired feature engineering

Contact Info

Product

Resources

About