Topic 1 Question 344
A machine learning (ML) engineer is creating a binary classification model. The ML engineer will use the model in a highly sensitive environment.
There is no cost associated with missing a positive label. However, the cost of making a false positive inference is extremely high.
What is the most important metric to optimize the model for in this scenario?
Accuracy
Precision
Recall
F1
ユーザの投票
コメント(3)
- 正解だと思う選択肢: B
from Copilot - In this scenario, the most important metric to optimize for is precision. Precision measures the proportion of true positive predictions among all positive predictions made by the model. Since the cost of making a false positive inference is extremely high, optimizing for precision will help minimize the number of false positives
👍 1MultiCloudIronMan2024/09/22 - 正解だと思う選択肢: B
In this scenario, the most important metric to optimize for is precision. Precision measures the proportion of true positive predictions among all positive predictions made by the model. Since the cost of making a false positive inference is extremely high, optimizing for precision will help minimize the number of false positives
👍 17f1fe732024/10/26 when FP cost is higher and important = precision when FN cost is higher and important = recall
👍 1spinatram2024/11/02
シャッフルモード