顶夸克标记器的机器学习全景
高能物理 - 唯象学
2019-07-31 v3
摘要
基于识别 boosted 强子衰变顶夸克这一既定任务，我们比较了广泛的现代机器学习方法。与大多数既定方法不同，它们依赖低级输入，例如量能器输出。尽管它们的网络架构差异巨大，但性能却较为相近。总体而言，我们发现这些新方法极其强大且非常有趣。
关键词
引用
@article{arxiv.1902.09914,
title = {The Machine Learning Landscape of Top Taggers},
author = {G. Kasieczka and T. Plehn and A. Butter and K. Cranmer and D. Debnath and B. M. Dillon and M. Fairbairn and D. A. Faroughy and W. Fedorko and C. Gay and L. Gouskos and J. F. Kamenik and P. T. Komiske and S. Leiss and A. Lister and S. Macaluso and E. M. Metodiev and L. Moore and B. Nachman and K. Nordstrom and J. Pearkes and H. Qu and Y. Rath and M. Rieger and D. Shih and J. M. Thompson and S. Varma},
journal= {arXiv preprint arXiv:1902.09914},
year = {2019}
}
备注
Yet another tagger included!