提出STRIDE框架，实现表格数据可解释性高效分解与交互分析。
STRIDE: Subset-Free Functional Decomposition for XAI in Tabular Settings
- 通过核空间正交分解，无需枚举子集即可解析特征作用
- 在10个数据集上比TreeSHAP快3倍，重构精度达R²=0.93
- 支持组件手术诊断，可量化单个交互项的影响
现有可解释AI方法多将特征影响简化为单一标量ϕ_i，仅回答‘什么重要’而忽略‘如何交互’。基于Shapley值的方法虽能捕捉交互，却面临指数级计算成本。本文提出STRIDE，一种无子集枚举、基于再生核希尔伯特空间（RKHS）的正交函数分解框架。在所研究的表格设置中，通过递归核中心化过程解析出各特征子集的函数分量f_S(x_S)。该方法模型无关且理论严谨，具有正交性和L²收敛性。在10个基准数据集（每组取10次随机种子）上，STRIDE相比TreeSHAP实现3.0倍中位速度提升，重建均方相关系数达R²=0.93。我们引入‘组件手术’诊断技术，可隔离并量化学习到的交互贡献；在加州房价数据集上，移除单一交互使测试R²从0.019降至0.027。
原文摘要 · Abstract (English)
Most explainable AI (XAI) frameworks are limited in their expressiveness, summarizing complex feature effects as single scalar values ϕ_i. This approach answers "what" features are important but fails to reveal "how" they interact. Furthermore, methods that attempt to capture interactions, like those based on Shapley values, often face an exponential computational cost. We present STRIDE, a scalable framework that addresses both limitations by reframing explanation as a subset-enumeration-free, orthogonal "functional decomposition" in a Reproducing Kernel Hilbert Space (RKHS). In the tabular setups we study, STRIDE analytically computes functional components f_S(x_S) via a recursive kernel-centering procedure. The approach is model-agnostic and theoretically grounded with results on orthogonality and L^2 convergence. In tabular benchmarks (10 datasets, median over 10 seeds), STRIDE attains a 3.0 times median speedup over TreeSHAP and a mean R^2=0.93 for reconstruction. We also introduce "component surgery", a diagnostic that isolates a learned interaction and quantifies its contribution; on California Housing, removing a single interaction reduces test R^2 from 0.019 to 0.027.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。