AWS Certification Exam AI Practitioner (Practice Questions) set9

問題 40 / 65

Which is the most appropriate reason why bias remains even if features indicating gender are removed from training?

Because deletion will inevitably reduce the model's accuracy.
Because the learning algorithm can recover gender.
Because there are variables that serve as proxies for gender, bias persists through those proxy variables.
Because removing it makes overfitting more likely.
Because the size of the training data becomes smaller.

当サイトでは、ユーザー体験の向上を目的としてCookieを使用しています。サイトの利用を継続することで、Cookieの使用に同意したものとみなされます。