NeuroRule：通过规则集进化使黑箱神经网络可解释
NeuroRule: Making Black-Box Neural Networks Explainable through Rule-set Evolution
- The University of Texas at Austin（德克萨斯大学奥斯汀分校）
- Cognizant AI Lab（高知特人工智能实验室）
机构由 AI 辅助整理，请以论文原文为准。
AI总结：
本文提出NeuroRule框架，利用进化算法将黑箱神经网络蒸馏为简洁可解释的规则集，无需原始训练数据，兼顾性能与可解释性，适用于高信任度场景。
AI中文摘要：
高容量神经网络模型在各种分类任务中已取得最先进的性能，然而它们常常作为黑箱模型运作，缺乏关键决策所需的透明度。这种不透明性造成了性能与可解释性之间持续存在的权衡。本文提出了一种解决这一差距的方案：NeuroRule知识蒸馏框架，该框架从神经网络模型中生成可解释的规则集。NeuroRule采用EVOTER规则集进化基础设施，将神经网络视为进化过程的目标，将其性能提炼为简洁的命题逻辑表达式集合。本文有三项主要贡献：（1）一种将黑箱神经网络模型蒸馏为显式规则集模型的进化方法；（2）一种通过将简洁性目标纳入进化过程来增强规则集可解释性的方法；（3）证明即使无法访问原始神经网络训练数据，蒸馏仍然可行。因此，本文确立了黑箱神经网络模型可以被赋予可解释性，从而在信任至关重要的现实应用中发挥作用。
英文摘要：
High-capacity neural network models have achieved state-of-the-art performance across diverse classification tasks, yet they frequently operate as black-box models, lacking the transparency necessary for critical decision-making. Such opacity creates a persistent trade-off between performance and explainability. This paper proposes a solution to address this gap: the NeuroRule knowledge distillation framework that results in explainable rule-sets from neural network models. NeuroRule adapts the EVOTER rule-set evolution infrastructure to treat neural networks as targets for the evolution process, distilling their performance into concise sets of propositional logic expressions. There are three primary contributions: (1) an evolutionary method for distilling black-box neural network models into explicit rule-set models; (2) a method for making rule sets more explainable by including a conciseness objective to evolution; and (3) a demonstration that the distillation is viable even without access to the original neural network training data. The paper thus establishes that black-box neural network models can be made explainable and therefore useful in real-world applications where trustworthiness is paramount.
