Interpretable boosted-decision-tree analysis for the Majorana Demonstrator
Interpretable boosted-decision-tree analysis for the Majorana Demonstrator
复制标题
马约拉纳演示器的可解释提升决策树分析
DOI：
10.1103/physrevc.107.014321
复制
发表时间：
2023
影响因子：
3.1
通讯作者：
Caldwell, T. S.
中科院分区：
文献类型：
--
作者：
Arnquist, I. J.;Avignone, F. T.;Barabash, A. S.;Barton, C. J.;Bhimani, K. H.;Blalock, E.;Bos, B.;Busch, M.;Buuck, M.;Caldwell, T. S.
Majorana Demonstrator是一个领先的实验，用高纯锗（HPGe）探测器寻找无中微子双β衰变。机器学习提供了一种新的方法来最大限度地提高这些检测器提供的信息量，但与传统分析相比，数据驱动的性质使其难以解释。可解释性研究揭示了机器的决策逻辑，使我们能够从机器学习中反馈到传统分析中。在这项工作中，我们提出了第一个机器学习分析的数据从马约拉纳演示;这也是第一个可解释的机器学习分析的任何锗探测器实验。训练两个梯度提升决策树模型从数据中学习，并进行基于博弈论的模型可解释性研究，以了解分类能力的来源。通过从数据中学习，该分析识别重建参数之间的相关性，以进一步提高背景抑制性能。通过向机器学习，这种分析揭示了新的背景类别对标准Majorana分析的重要性。该模型与下一代锗探测器实验（如LEGEND）高度兼容，因为它可以在大量探测器上同时训练。
TheMajorana Demonstratoris a leading experiment searching for neutrinoless double-beta decay with high purity germanium (HPGe) detectors. Machine learning provides a new way to maximize the amount of information provided by these detectors, but the data-driven nature makes it less interpretable compared to traditional analysis. An interpretability study reveals the machine's decision-making logic, allowing us to learn from the machine to feed back to the traditional analysis. In this work, we present the first machine learning analysis of the data from theMajorana Demonstrator; this is also the first interpretable machine learning analysis of any germanium detector experiment. Two gradient boosted decision tree models are trained to learn from the data, and a game-theory-based model interpretability study is conducted to understand the origin of the classification power. By learning from data, this analysis recognizes the correlations among reconstruction parameters to further enhance the background rejection performance. By learning from the machine, this analysis reveals the importance of new background categories to reciprocally benefit the standardMajoranaanalysis. This model is highly compatible with next-generation germanium detector experiments like LEGEND since it can be simultaneously trained on a large number of detectors.