Skip to main content
QUICK REVIEW

[论文解读] Generative Machine Learning for Detector Response Modeling with a Conditional Normalizing Flow

Allison Xu, S. Han|arXiv (Cornell University)|Mar 17, 2023
Particle physics theoretical and experimental studies被引用 10
一句话总结

论文开发了一个条件正态流(CNF)模型,用来模拟 Higgs 到二光子事件的探测器响应,捕捉相关性和非对称效应,以替代成本高昂的蒙特卡洛模拟。

ABSTRACT

In this paper, we explore the potential of generative machine learning models as an alternative to the computationally expensive Monte Carlo (MC) simulations commonly used by the Large Hadron Collider (LHC) experiments. Our objective is to develop a generative model capable of efficiently simulating detector responses for specific particle observables, focusing on the correlations between detector responses of different particles in the same event and accommodating asymmetric detector responses. We present a conditional normalizing flow model (CNF) based on a chain of Masked Autoregressive Flows, which effectively incorporates conditional variables and models high-dimensional density distributions. We assess the performance of the \cnf model using a simulated sample of Higgs boson decaying to diphoton events at the LHC. We create reconstruction-level observables using a smearing technique. We show that conditional normalizing flows can accurately model complex detector responses and their correlation. This method can potentially reduce the computational burden associated with generating large numbers of simulated events while ensuring that the generated events meet the requirements for data analyses.

研究动机与目标

  • 动机:用生成方法替代对特定观测量的昂贵 MC 探测器仿真。
  • 学习以事件变量和粒子运动学为条件的探测器响应。
  • 建模同一事件中多个粒子的探测器响应之间的相关性。
  • 处理超越简单 smeared 技术的非对称探测器响应分布。
  • 展示对 Higgs 到二光子分析的适用性,并有潜力扩展到其他探测器和观测量。

提出的方法

  • 使用由带置换的 Masked Autoregressive Flows (MAF) 链构成的条件正则流。
  • 将 CNF 以 pileup 和真实粒子运动学为条件,以建模高维探测器响应分布。
  • 将六个探测器响应变量(ΔX)用于两光子的 E_T、η 和 φ,作为一个 6D 目标,配以 6D 基密度。
  • 将输出缩放到 [-1,1],并应用 tanh 双射以保持结果在范围内。
  • 使用 Adam 进行 500 个 epoch 的训练,采用幂律学习率衰减,并监控六个探测器变量的平均 Wasserstein 距离。
  • 将数据分为 80/10/10 的训练/验证/测试,按最小平均 WD 选择最佳模型。
Figure 1 : The mean Wasserstein Distance (orange) and the minimum Wasserstein distance (blue) as a function of the training epochs for the baseline scenario. These quantities were evaluated on the validation sample.
Figure 1 : The mean Wasserstein Distance (orange) and the minimum Wasserstein distance (blue) as a function of the training epochs for the baseline scenario. These quantities were evaluated on the validation sample.

实验结果

研究问题

  • RQ1CNF 是否能学习以事件变量为条件的两光子探测器响应的完整联合分布?
  • RQ2CNF 是否能重现光子探测器响应之间的相关性并处理非对称探测器效应?
  • RQ3CNF 在逼近目标探测器分辨率以及 diphoton 质量和横向动量分布方面的表现如何?
  • RQ4在复杂依赖场景中,与传统 smeared 方法相比的性能含义是什么?

主要发现

  • 基线情景显示目标与 CNF 学得的探测器分辨率之间有很好的一致性(<5% 偏差)。
  • CNF 在探测器层面再现光子动量变量(E_T、η、φ)的分布,吻合良好。
  • 在相关性情景中,CNF 能准确捕捉两光子之间在 ρ = 0.5 与 ρ = 1.0 的相关性。
  • 非对称探测器响应情景中,CNF 能重现探测器响应分布的非对称尾部。
  • 与 CNF 一致的 diphoton 不变量质量和 diphoton 系统 p_T 分布在统计精度内。
Figure 2 : Target and generated photon resolutions $\sigma$ for photon kinematic variables $E_{\text{T}}$ , $\eta$ , and $\phi$ . The resolutions are shown as functions of the true values of photon $E_{\text{T}}$ and $\eta$ , and the event pile-up $\mu$ . The blue (orange) entries represent the targ
Figure 2 : Target and generated photon resolutions $\sigma$ for photon kinematic variables $E_{\text{T}}$ , $\eta$ , and $\phi$ . The resolutions are shown as functions of the true values of photon $E_{\text{T}}$ and $\eta$ , and the event pile-up $\mu$ . The blue (orange) entries represent the targ

更好的研究,从现在开始

从论文设计到论文写作,大幅缩短您的研究时间。

无需绑定信用卡

本解读由 AI 生成,并经人工编辑审核。