提出可解释的模型漂移检测方法，兼顾准确与原因分析。
Interpretable Model Drift Detection
- 基于特征交互感知的假设检验框架，从风险角度建模漂移
- 在真实数据集上优于现有可解释方法，接近黑箱顶尖水平
- 能定位对漂移敏感的关键特征，适合需理解变化原因的场景
现实世界中的数据分布会随时间演变，导致训练好的机器学习模型逐渐过时，这种现象称为模型漂移。识别漂移有助于保持模型准确性，并发现输入特征与输出变量间关系的变化。现有方法多仅关注漂移检测，缺乏可解释性。本文从风险视角出发，提出一种具备特征交互感知的可解释漂移检测框架，具有测试功效保障。该方法通用性强，适用于分类与回归任务。在多个标准漂移检测数据集上的实验表明，本方法在可解释性方面优于现有方法（尤其在真实数据集上），性能与当前最优黑箱方法相当。我们还通过定量与定性分析验证了其可解释性，包括对USENET2数据集的案例研究，发现该方法更聚焦于模型和漂移敏感的特征，相较基线可解释检测器更具针对性。
原文摘要 · Abstract (English)
Data in the real world often has an evolving distribution. Thus, machine learning models trained on such data get outdated over time. This phenomenon is called model drift. Knowledge of this drift serves two purposes: (i) Retain an accurate model and (ii) Discovery of knowledge or insights about change in the relationship between input features and output variable w.r.t. the model. Most existing works focus only on detecting model drift but offer no interpretability. In this work, we take a principled approach to study the problem of interpretable model drift detection from a risk perspective using a feature-interaction aware hypothesis testing framework, which enjoys guarantees on test power. The proposed framework is generic, i.e., it can be adapted to both classification and regression tasks. Experiments on several standard drift detection datasets show that our method is superior to existing interpretable methods (especially on real-world datasets) and on par with state-of-the-art black-box drift detection methods. We also quantitatively and qualitatively study the interpretability aspect including a case study on USENET2 dataset. We find our method focuses on model and drift sensitive features compared to baseline interpretable drift detectors.
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。