Objective 3.3 Carry Out Feature Engineering CERTIFIED-DATA-SCIENCE-PRACTITIONER Practice Questions (Page 10)

Questions 46–50

1. 46expert · hard

   A data scientist is performing feature selection on a dataset with 1,000 features and 5,000 rows. The data scientist is using statistical tests to evaluate each feature's association with the target. The goal is to control the expected proportion of false positives when selecting important features. Which method is most appropriate?

   • ARandom forest feature importance
   • BHigh-correlation filter
   • CLow-variance filter
   • DFalse discovery rate (FDR) control
   Reveal Solution Discussion Select an answer first
2. 47application · easy

   A data scientist is working with a dataset containing a 'PurchaseDate' column with timestamps. The goal is to build a model to predict customer churn. Which feature engineering step is most appropriate?

   • AExtract day of week, month, and year from the date
   • BUse the raw timestamp as a numeric feature
   • CDrop the date column
   • DConvert the date to a string and use it as a categorical feature
   Reveal Solution Discussion Select an answer first
4. 48application · easy

   A data scientist is building a decision tree model to predict house prices. The dataset includes an 'EducationLevel' feature with categories 'High School', 'Bachelor', 'Master', and 'PhD'. The categories have a natural order. Which encoding method is most appropriate?

   • ADummy encoding
   • BTarget encoding
   • CLabel encoding with values 0, 1, 2, 3
   • DOne-hot encoding
   Reveal Solution Discussion Select an answer first
5. 49foundation · easy

   A data scientist wants to encode a categorical feature using a numeral system with base 4. How many columns will be needed to represent 64 categories?

   • A4
   • B6
   • C3
   • D64
   Reveal Solution Discussion Select an answer first
6. 50foundation · easy

   What is the purpose of using a missing value ratio filter?

   • ATo convert missing values into binary indicators
   • BTo impute missing values with the mean
   • CTo remove features with a high proportion of missing values
   • DTo rank features by their importance
   Reveal Solution Discussion Select an answer first