You are working on a binary classification ML algorithm that detects whether an image of a classified scanned document contains a company's logo. In the dataset, 96% of examples don't have the logo, s

The correct answer is D. F1 score. With 96% negative examples, a naive model that always predicts 'no logo' achieves 96% accuracy, making accuracy useless. Precision alone (answer A) measures how often positive predictions are correct but ignores missed detections. Recall alone (answer B) measures coverage of actu

Submitted by ngozi_ng· Apr 18, 2026ML model development

Question

You are working on a binary classification ML algorithm that detects whether an image of a classified scanned document contains a company's logo. In the dataset, 96% of examples don't have the logo, so the dataset is very skewed. Which metric would give you the most confidence in your model?

Options

APrecision
BRecall
CRMSE
DF1 score

How the community answered

(47 responses)

A
21% (10)
B
6% (3)
C
9% (4)
D
64% (30)

Explanation

With 96% negative examples, a naive model that always predicts 'no logo' achieves 96% accuracy, making accuracy useless. Precision alone (answer A) measures how often positive predictions are correct but ignores missed detections. Recall alone (answer B) measures coverage of actual positives but ignores false alarms. RMSE (answer C) is a regression metric and does not apply here. The F1 score (answer D) is the harmonic mean of precision and recall, penalizing models that sacrifice one for the other. It is the standard metric for imbalanced binary classification because it forces the model to perform well on both dimensions simultaneously.

Topics

#Evaluation Metrics#Imbalanced Data#F1 Score#Binary Classification

Community Discussion

No community discussion yet for this question.

Full PROFESSIONAL-MACHINE-LEARNING-ENGINEER Practice