课程: Building Trustworthy AI Systems: Transparency, Explainability, and Control with ISO/IEC TR 24028

免费学习该课程！

今天就开通帐号，24,600 门业界名师课程任您挑！

Evaluation (Clauses 9.10.3-9.10.5)

Evaluation (Clauses 9.10.3-9.10.5)

课程: Building Trustworthy AI Systems: Transparency, Explainability, and Control with ISO/IEC TR 24028

开始一个月试用 购买团队方案

Evaluation (Clauses 9.10.3-9.10.5)

“

- [Instructor] When was the last time you went to a doctor for an annual health checkup? What characterizes the visit? Does your doctor ask you questions about your overall health and if you have been having issues with anything you'd like to talk about? Does your doctor then check your vitals and let you know if they fall within the normal range of performance? Things like blood pressure, heart rate, and metabolic measures. Your annual health check is really an evaluation of the robustness of your health. In the same manner, when we evaluate an AI system, we check its robustness. I already stated that robustness is the degree to which an AI system can maintain its intended level of performance despite potentially disruptive or widely varying environmental influences. Different metrics can be used to assess different AI use cases. Here are a few examples. Let's say an AI system is tasked with interpolation or making a best guess of known values between two known points. So the two…

内容