A data engineer is using Amazon SageMaker Data Wrangler to evaluate customer data for a behavior prediction model. To improve model performance they need to check for multicollinearity. Which steps require the least operational effort to detect multicollinearity? (Choose two.)
Choose an answer
Tap an option to check your answer.
Correct answer: Use SageMaker Data Wrangler diagnostic visualizations and run principal component analysis (PCA) and singular value decomposition (SVD) to obtain singular values., Use SageMaker Data Wrangler diagnostic visualizations and run LASSO to plot coefficient values from a LASSO model trained on the dataset..
Why this is the answer
The correct options leverage SageMaker Data Wrangler's built-in diagnostic visualizations for multicollinearity detection with minimal effort. PCA and SVD can identify multicollinearity by analyzing eigenvalues or singular values, where small values indicate highly correlated features. LASSO (Least Absolute Shrinkage and Selection Operator) is a regularization technique that can shrink the coefficients of correlated features towards zero, effectively highlighting multicollinearity by showing which features become redundant. Refitting and transforming with one-hot encoding is a preprocessing step, not a direct multicollinearity detection method. Quick Model provides feature importance but doesn't directly diagnose multicollinearity. Min Max Scaler normalizes data but doesn't detect multicollinearity.
Pass your exam — without the endless answer hunt
Get every verified question and explanation for this exam in one place, and save hours of prep. 1,000+ certifications · 20+ languages · free to start.
Pass your exam faster → No card needed