Objective 3.2 Preprocess Data CERTIFIED-DATA-SCIENCE-PRACTITIONER Practice Questions (Page 3)

Questions 11–15

1. 11foundation · easy

   Which scaling method is most robust to outliers?

   • ARobust scaling
   • BMin-max scaling
   • CZ-score standardization
   • DLog transformation
   Reveal Solution Discussion Select an answer first
2. 12application · medium

   A data scientist is preparing features for a neural network. The feature 'age' ranges from 18 to 90, and 'income' ranges from 20,000 to 200,000. The neural network uses gradient descent and requires features to be on a similar scale. The data scientist wants to preserve the original distribution shape and bound the values between 0 and 1. What should the data scientist do?

   • AApply log transformation to income and min-max scaling to age.
   • BApply robust scaling to both features.
   • CApply standardization (z-score) to both features.
   • DApply min-max scaling to both features.
   Reveal Solution Discussion Select an answer first
4. 13foundation · easy

   Which type of missing data occurs when the probability of missingness depends on the value of the missing variable itself?

   • AMNAR
   • BMAR
   • CMNR
   • DMCAR
   Reveal Solution Discussion Select an answer first
5. 14application · medium

   A data scientist is working on a dataset where the 'age' column has missing values. The missingness is completely random (MCAR). The dataset is large (100,000 rows) and the 'age' column is skewed to the right. The data scientist wants to impute the missing values in a way that minimally distorts the distribution of 'age'. What should the data scientist do?

   • AImpute with the mean age.
   • BImpute with the median age.
   • CDelete all rows with missing age.
   • DImpute with the mode age.
   Reveal Solution Discussion Select an answer first
6. 15expert · hard

   A data scientist is imputing missing values in a dataset with a mix of continuous and categorical features. The missingness is MAR. The data scientist wants to use an imputation method that accounts for relationships between features. The dataset is moderately large (50,000 rows). What is the most appropriate imputation approach?

   • ADelete rows with any missing values.
   • BUse multiple imputation by chained equations (MICE).
   • CImpute each feature with its median.
   • DImpute each feature with its mode.
   Reveal Solution Discussion Select an answer first