Databricks Machine Learning Associate Question 596
Single answerYou are working on a binary classification problem in Databricks and have trained multiple models. To evaluate and compare the models, you examine their metrics. One of the models has a high F1 score but a low ROC AUC score. What does this indicate about the model's performance?
- A
The model has a good balance between precision and recall but struggles with overall ranking of positive and negative classes.
- B
The model is overfitting and has poor generalization to unseen data.
- C
The model performs well at ranking positive and negative classes but has poor precision and recall.
- D
The model is underfitting and fails to capture relevant patterns in the data.
Show answer and explanation
Correct answer: A
Explanation
The F1 score measures the harmonic mean of precision and recall, which is important for imbalanced datasets or when false positives and false negatives have similar costs. Meanwhile, the ROC AUC score evaluates the model's ability to rank predictions correctly. A high F1 but low ROC AUC score indicates that the model predicts well within its decision threshold but struggles with ranking positive and negative classes across various thresholds.
- A. Correct.
This is correct. A high F1 score indicates a good balance between precision and recall, while a low ROC AUC score suggests the model struggles to rank positive and negative classes correctly.
- B. Incorrect.
This is incorrect. A high F1 score suggests the model is not overfitting, as it effectively balances precision and recall.
- C. Incorrect.
This is incorrect. A low ROC AUC score indicates a poor ranking capability, but a high F1 score disproves poor precision and recall.
- D. Incorrect.
This is incorrect. Underfitting would result in both low F1 and ROC AUC scores, which is not the case here.