AWS Certification Exam AI Practitioner (Practice Questions) set16
通常
試験時間: 00:00問題時間: 00:00
問題一覧
問題 53 / 65
In the production environment, it is suspected that the model's output probabilities have become unreliable. Which metric is most appropriate for evaluating the calibration of predicted probabilities?
①Brier score
②AUC (area under the ROC curve)
③Accuracy (accuracy)
④F1 score
⑤Summary of the confusion matrix