FLEX:基于 Fisher LDA 的多标签分类器全局解释方法
Right Labels, Wrong Reasons: Signed Feature Attribution for Multi-Label Explainability
多标签分类器经常蒙对答案但理由不对。这篇论文用 Fisher LDA 造了个全局解释器,一次输出所有标签的带符号特征归因,还能标记哪些标签可能是蒙对的。做分类模型审计和可解释性研究的人值得跑一下代码。
arXiv 论文提出 FLEX(Fisher LDA Explainer),一种针对多标签分类器的全局代理解释方法。它在完整的二值标签矩阵上拟合单个多标签 Fisher LDA,一次生成每个标签的带符号特征归因列表,所有标签共享同一坐标系。
方法上,FLEX 跨标签池化协方差,在大多数数据集上优于逐标签方案,且在标签数量多、正样本稀少的场景中优势最大。其保真度-复杂度 Pareto 曲线单调递增,意味着增加特征预算可稳定提升解释保真度;在中等特征预算下,样本内保真度可持平或超过 Binary Relevance 基线。
论文还指出,各判别权重向量的大小可约束对应标签的可分性,用来识别"标签正确但理由错误"的情况。实验显示不同分类器家族的预测结论大体一致,但在无主导特征时依赖的特征子集并不相同。
Right Labels, Wrong Reasons: Signed Feature Attribution for Multi-Label Explainability
When a multi-label classifier assigns the right labels, it may do so for the wrong reasons. Diffuse feature combinations may vary across classifier families and lack a dominant feature. This paper proposes FLEX (Fisher LDA Explainer), a global surrogate that fits a single direct multi-label Fisher LDA over the full binary label matrix, producing signed feature attribution lists for each label in a single coherent coordinate system. Across the evaluated benchmarks and classifier families, FLEX reproduces the black box's predictions with consistently high fidelity. Pooling covariance across labels outperforms a per-label alternative on most datasets, with the advantage growing largest where labels are numerous and positive examples are scarce. FLEX's fidelity-complexity Pareto curve is monotonically increasing, giving practitioners a predictable path to higher-fidelity explanations as more features are used, and its in-sample fidelity matches or leads Binary Relevance alternatives once a modest feature budget is allowed. The magnitude of each discriminant weight vector bounds its label's separability, flagging labels that may be predicted correctly for the wrong reasons. The evaluated classifiers largely agree on which predictions to make, but reach those decisions through different feature subsets when no single feature dominates.
本条情报归属 AITOP「论文」核心追踪栏目。在跨区域产业观察坐标中,该动态直接关联「AI 开发者工具链、Coding Agent 与工程化实践」研究主线,反映出当前技术演进与商业落地的最新风向。涉及核心实体与技术议题包括:Fisher LDA、多标签分类、可解释性、特征归因,读者可通过下方标签进入对应专题页查阅历史脉络与关联对比。
信源核验说明:本文原始线索采编自「arXiv cs.LG」(@Kenton White),经 AITOP 多维评分管线结构化提炼。