Prepare and Clean Data SNOWPRO-ADVANCED-DATA-SCIENTIST Practice Questions (Page 5)

Questions 21–22

1. 21application · medium

   A data scientist is building a decision tree classifier. The dataset includes a categorical feature 'Region' with 12 distinct values, some with very low frequency. The scientist wants to avoid high-dimensionality and preserve the tree's interpretability. Which encoding strategy is most appropriate?

   • AGroup rare categories into an 'Other' category, then apply one-hot encoding.
   • BApply one-hot encoding to all 12 categories.
   • CApply label encoding to the categories.
   • DUse target encoding based on the mean of the target variable.
   Reveal Solution Discussion Select an answer first
2. 22application · medium

   A data scientist receives a dataset from an external vendor. The 'date' column is in the format 'YYYYMMDD' as an integer, and the 'price' column has some values with leading/trailing spaces. The scientist needs to prepare the data for time-series analysis. Which sequence of cleaning steps is most appropriate?

   • AConvert 'date' to string and 'price' to integer.
   • BStrip whitespace from 'date' and convert 'price' to numeric.
   • CConvert the 'date' integer to a datetime type, strip whitespace from 'price', and convert 'price' to a numeric type.
   • DConvert 'price' to numeric first, then convert 'date' to datetime.
   Reveal Solution Discussion Select an answer first