학술논문
Developing a Bubble Chamber Particle Discriminator Using Semi-Supervised Learning
Document Type
Working Paper
Author
Matusch, B.; Amole, C.; Ardid, M.; Arnquist, I. J.; Asner, D. M.; Baxter, D.; Behnke, E.; Bressler, M.; Broerman, B.; Cao, G.; Chen, C. J.; Chowdhury, U.; Clark, K.; Collar, J. I.; Cooper, P. S.; Coutu, C. B.; Cowles, C.; Crisler, M.; Crowder, G.; Cruz-Venegas, N. A.; Dahl, C. E.; Das, M.; Fallows, S.; Farine, J.; Felis, I.; Filgas, R.; Girard, F.; Giroux, G.; Hall, J.; Hardy, C.; Harris, O.; Hillier, T.; Hoppe, E. W.; Jackson, C. M.; Jin, M.; Klopfenstein, L.; Krauss, C. B.; Laurin, M.; Lawson, I.; Leblanc, A.; Levine, I.; Licciardi, C.; Lippincott, W. H.; Loer, B.; Mamedov, F.; Mitra, P.; Moore, C.; Nania, T.; Neilson, R.; Noble, A. J.; Oedekerk, P.; Ortega, A.; Piro, M. -C.; Plante, A.; Podviyanuk, R.; Priya, S.; Robinson, A. E.; Sahoo, S.; Scallon, O.; Seth, S.; Sonnenschein, A.; Starinski, N.; Štekl, I.; Sullivan, T.; Tardif, F.; Vázquez-Jáuregui, E.; Walkowski, N.; Weima, E.; Wichoski, U.; Wierman, K.; Yan, Y.; Zacek, V.; Zhang, J.
Source
Subject
Language
Abstract
The identification of non-signal events is a major hurdle to overcome for bubble chamber dark matter experiments such as PICO-60. The current practice of manually developing a discriminator function to eliminate background events is difficult when available calibration data is frequently impure and present only in small quantities. In this study, several different discriminator input/preprocessing formats and neural network architectures are applied to the task. First, they are optimized in a supervised learning context. Next, two novel semi-supervised learning algorithms are trained, and found to replicate the Acoustic Parameter (AP) discriminator previously used in PICO-60 with a mean of 97% accuracy.
Comment: 27 pages, 10 figures
Comment: 27 pages, 10 figures