机器学习模型的模型无关解释方法的一般缺陷
General Pitfalls of Model-Agnostic Interpretation Methods for Machine Learning Models
摘要：
越来越多的用于机器学习（ML）模型的模型不可知解释技术，例如部分依赖图（PDP）、置换特征重要性（PFI）和Shapley值，提供了有洞察力的模型解释，但如果应用不当，可能会导致错误的结论。我们强调了ML模型解释的许多常见陷阱，例如在错误的上下文中使用解释技术，解释不能很好地泛化的模型，忽略高维环境中的特征依赖性，相互作用，不确定性估计和问题，或者做出不合理的因果解释，并通过示例进行说明。我们专注于描述平均模型行为的全局方法的陷阱，但许多陷阱也适用于解释单个预测的局部方法。我们的论文通过提高对陷阱的认识和确定正确模型解释的解决方案来解决ML从业者，但也通过讨论进一步研究的开放问题来解决ML研究人员。
英文摘要：
An increasing number of model-agnostic interpretation techniques for machine learning (ML) models such as partial dependence plots (PDP), permutation feature importance (PFI) and Shapley values provide insightful model interpretations, but can lead to wrong conclusions if applied incorrectly. We highlight many general pitfalls of ML model interpretation, such as using interpretation techniques in the wrong context, interpreting models that do not generalize well, ignoring feature dependencies, interactions, uncertainty estimates and issues in high-dimensional settings, or making unjustified causal interpretations, and illustrate them with examples. We focus on pitfalls for global methods that describe the average model behavior, but many pitfalls also apply to local methods that explain individual predictions. Our paper addresses ML practitioners by raising awareness of pitfalls and identifying solutions for correct model interpretation, but also addresses ML researchers by discussing open issues for further research.
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。
