Objective 3.3 Carry Out Feature Engineering CERTIFIED-DATA-SCIENCE-PRACTITIONER Practice Questions (Page 11)

Questions 51–54

1. 51foundation · easy

   What is the purpose of splitting a feature like 'full_name' into 'first_name' and 'last_name'?

   • ATo reduce the number of features in the dataset
   • BTo create more granular features that may improve model performance
   • CTo handle missing values in the original feature
   • DTo convert text data into numeric data
   Reveal Solution Discussion Select an answer first
2. 52foundation · easy

   A dataset has a column 'full_name' with values like 'John Michael Smith'. The data scientist wants to create separate columns for first, middle, and last names. Which feature engineering technique should they use?

   • ASplit names
   • BTarget encoding
   • COne-hot encoding
   • DPCA
   Reveal Solution Discussion Select an answer first
4. 53foundation · medium

   A data science team is building a churn prediction model with 120 candidate features. They want to reduce the feature set by starting with all 120 features and, at each iteration, removing the feature whose removal degrades model performance the least. Which feature selection technique does this describe?

   • AForward feature selection
   • BLow-variance filter
   • CBackward feature elimination
   • DRandom forest feature importance
   Reveal Solution Discussion Select an answer first
5. 54foundation · easy

   A dataset contains movie titles like 'Inception (2010)'. The data scientist wants to extract the release year as a separate feature. Which feature engineering technique should they use?

   • APCA
   • BTarget encoding
   • CExtract year from title
   • DOne-hot encoding
   Reveal Solution Discussion Select an answer first