MLA-C01 Question 227
Select 4You are using Amazon SageMaker Clarify to detect potential bias in your machine learning model and training data. Which of the following metrics provided by SageMaker Clarify can help you assess bias and explainability in your dataset and model predictions?
- A
Demographic Parity Difference
- B
Class Imbalance Ratio
- C
Feature Importance
- D
Conditional Demographic Disparity
- E
Shapley Values
Show answer and explanation
Correct answers: A, C, D, E
Explanation
SageMaker Clarify provides metrics to assess both bias and explainability in machine learning datasets and models. Bias metrics such as Demographic Parity Difference and Conditional Demographic Disparity help identify potential bias in data or predictions across demographic groups, while explainability metrics like Feature Importance and Shapley Values provide insights into how features influence model predictions. Class Imbalance Ratio, while important for understanding data distribution, is not a metric provided by SageMaker Clarify.
- A. Correct.
Demographic Parity Difference is a bias metric provided by SageMaker Clarify that measures the difference in positive outcomes across different demographic groups.
- B. Incorrect.
Class Imbalance Ratio is not a bias or explainability metric provided by SageMaker Clarify. It is a general characteristic of datasets but not specific to Clarify.
- C. Correct.
Feature Importance is provided by SageMaker Clarify and is used to understand the relative importance of input features in model predictions.
- D. Correct.
Conditional Demographic Disparity is a bias metric provided by SageMaker Clarify that evaluates bias conditioned on certain features.
- E. Correct.
Shapley Values are used by SageMaker Clarify for explainability of model predictions, helping to understand how individual features contribute to a specific prediction.