BayesNAM：利用不一致性实现可靠解释
BayesNAM: Leveraging Inconsistency for Reliable Explanations
AI总结：
本文指出神经加性模型（NAM）产生的不一致解释具有内在价值，提出BayesNAM框架，结合贝叶斯神经网络与特征dropout来捕捉不一致性，从而提供更可靠的解释并揭示数据与模型结构的潜在问题。
摘要：
神经加性模型（NAM）是一种最近提出的可解释人工智能（XAI）方法，利用基于神经网络的架构。鉴于神经网络的优势，NAM在提供高模型性能的同时，为其预测提供直观的解释。在本文中，我们分析了一个关键却被忽视的现象：即使使用相同的架构和数据集，NAM也经常产生不一致的解释。传统上，这种不一致性被视为需要解决的问题。然而，我们反而认为，在给定的数据模型内，这些不一致性可以提供有价值的解释。通过一个简单的理论框架，我们证明这些不一致性并非单纯的伪影，而是在具有多个重要特征的数据集中自然产生的。为了有效利用这些信息，我们引入了一个新框架：贝叶斯神经加性模型（BayesNAM），它集成了贝叶斯神经网络和特征dropout，理论证明表明特征dropout能有效捕捉模型不一致性。我们的实验表明，BayesNAM能有效揭示数据不足或模型结构限制等潜在问题，提供更可靠的解释和潜在补救措施。
英文摘要：
Neural additive model (NAM) is a recently proposed explainable artificial intelligence (XAI) method that utilizes neural network-based architectures. Given the advantages of neural networks, NAMs provide intuitive explanations for their predictions with high model performance. In this paper, we analyze a critical yet overlooked phenomenon: NAMs often produce inconsistent explanations, even when using the same architecture and dataset. Traditionally, such inconsistencies have been viewed as issues to be resolved. However, we argue instead that these inconsistencies can provide valuable explanations within the given data model. Through a simple theoretical framework, we demonstrate that these inconsistencies are not mere artifacts but emerge naturally in datasets with multiple important features. To effectively leverage this information, we introduce a novel framework, Bayesian Neural Additive Model (BayesNAM), which integrates Bayesian neural networks and feature dropout, with theoretical proof demonstrating that feature dropout effectively captures model inconsistencies. Our experiments demonstrate that BayesNAM effectively reveals potential problems such as insufficient data or structural limitations of the model, providing more reliable explanations and potential remedies.
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。
