arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2110.08508hep-phhep-exphysics.data-an
收藏

改进变分自动编码器用于LHC的新物理探测

Improving Variational Autoencoders for New Physics Detection at the LHC with Normalizing Flows

Pratik Jawahar, Thea Aarrestad, Nadezda Chernyavskaya, Maurizio Pierini, Kinga A. Wozniak, Jennifer Ngadiuba, Javier Duarte, Steven Tsan

摘要:

我们研究如何提高新的物理检测策略,利用变分自动编码器和规范化流程的异常检测在大型强子对撞机。作为一个工作示例,我们考虑DarkMachines挑战数据集。我们展示了不同的设计选择(例如,事件表示、异常分数定义、网络体系结构)影响特定基准新物理模型的结果。一旦建立了基线,我们讨论了如何通过利用变分自动编码器的潜在空间中的归一化流层来提高异常检测的准确性。

英文摘要:

We investigate how to improve new physics detection strategies exploiting variational autoencoders and normalizing flows for anomaly detection at the Large Hadron Collider. As a working example, we consider the DarkMachines challenge dataset. We show how different design choices (e.g., event representations, anomaly score definitions, network architectures) affect the result on specific benchmark new physics models. Once a baseline is established, we discuss how to improve the anomaly detection accuracy by exploiting normalizing flow layers in the latent space of the variational autoencoder.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑