笔记暂时采用中英夹杂的模式,会不断polish,也欢迎读者的建议
Explainable Fuzzy Systems: Paving the Way from Interpretable Fuzzy Systems to Explainable AI Systems. (英文书名)
Chapter 1: Toward Explainable Artificial Intelligence Through Fuzzy Systems.
In critical contexts, an XAI (Explainable AI) must be able to explain its choices in order to guarantee that some safety concerns are met.
此外,随着机器学习系统复杂性的增加,解释甚至有助于帮助用户调试或更好地理解系统行为。
Lipton (2018) 提出来两个专有名词
1. Transparency: a quality connected to understanding the working mechanisms of a model. The quality of transparency can be further refined in simulatability (i.e. ), decomposability (i.e. ), algorithmic transparency (i.e. the)
2. Post-hoc interpretability.