MLS-C01 Question #329: Real Exam Question with Answer & Explanation

Question

A chemical company has developed several machine learning (ML) solutions to identify chemical process abnormalities. The time series values of independent variables and the labels are available for the past 2 years and are sufficient to accurately model the problem. The regular operation label is marked as 0 The abnormal operation label is marked as 1. Process abnormalities have a significant negative effect on the company's profits. The company must avoid these abnormalities. Which metrics will indicate an ML solution that will provide the GREATEST probability of detecting an abnormality?

Options

• APrecision = 0.91
• BPrecision = 0.61
• CPrecision = 0.7
• DPrecision = 0.98

Explanation

To maximize the detection of critical process abnormalities when they have a significant negative effect, a solution should prioritize recall, which often correlates with accepting a lower precision value.

Common mistakes.

• A. Precision = 0.91 indicates a model that is more accurate when it predicts an abnormality, but it doesn't guarantee a high recall for detecting all abnormalities.
• C. Precision = 0.7 is a moderate precision value; a higher precision doesn't prioritize catching all abnormalities over avoiding false alarms.
• D. Precision = 0.98 indicates a very high confidence in positive predictions but suggests the model is highly conservative, which could lead to missing many actual abnormalities (low recall), contradicting the goal of detecting the 'GREATEST probability of detecting an abnormality'.

Concept tested. Machine learning evaluation metrics (Recall vs Precision)

Reference. https://docs.aws.amazon.com/sagemaker/latest/dg/clarify-concepts-performance-metrics.html

No community discussion yet for this question.