Handle Missing and Categorical Data SNOWPRO-ADVANCED-DATA-SCIENTIST Practice Questions (Page 3)

Questions 11–15

1. 11application · medium

   A data scientist is preparing data for a gradient boosting model to predict customer churn. The dataset has a 'zip_code' feature with 5,000 unique values. The team wants to capture the predictive power of the zip code without creating excessive dimensionality. What is the most appropriate approach?

   • AApply label encoding.
   • BApply target encoding, using the mean of the target variable for each zip code.
   • COne-hot encode 'zip_code'.
   • DDrop the 'zip_code' feature entirely.
   Reveal Solution Discussion Select an answer first
2. 12expert · hard

   A data scientist is preparing data for a gradient boosting model (XGBoost). The dataset has a categorical feature 'education_level' with missing values, and a continuous feature 'age' with missing values. The missingness is MCAR. The team wants to minimize preprocessing bias and leverage the model's native capabilities. What is the most efficient approach?

   • AUse XGBoost's built-in missing value handling by leaving the missing values as NaN and encoding 'education_level' with label encoding.
   • BDrop all rows with any missing values to ensure a clean dataset.
   • CImpute 'age' with the mean and 'education_level' with the mode, then one-hot encode.
   • DPerform multiple imputation using MICE, then train the XGBoost model on each imputed dataset and pool the results.
   Reveal Solution Discussion Select an answer first
4. 13foundation · easy

   A data scientist is imputing missing values in a numeric feature that is heavily skewed, with a few extreme outliers. Which imputation method is most appropriate to minimize the impact of outliers?

   • AMean imputation
   • BMode imputation
   • CConstant imputation with zero
   • DMedian imputation
   Reveal Solution Discussion Select an answer first
5. 14foundation · easy

   A categorical feature represents education level with values 'High School', 'Bachelor's', 'Master's', and 'PhD'. Which encoding method is most appropriate to preserve the natural order of these categories?

   • ABinary encoding
   • BLabel encoding
   • COne-hot encoding
   • DOrdinal encoding
   Reveal Solution Discussion Select an answer first
6. 15application · medium

   A data scientist is preparing a dataset for a linear regression model. The 'number_of_rooms' feature is a count variable with a few missing values. The distribution is roughly normal with no extreme outliers. The missingness is MCAR. What is the most appropriate imputation method?

   • AKNN imputation
   • BMean imputation
   • CMode imputation
   • DMedian imputation
   Reveal Solution Discussion Select an answer first