Engineer Features for Models SNOWPRO-ADVANCED-DATA-SCIENTIST Practice Questions (Page 2)

Questions 6–10

1. 6foundation · easy

   What is the purpose of automated feature engineering tools?

   • ATo impute missing values in the dataset
   • BTo visualize the dataset
   • CTo systematically generate candidate features from raw data
   • DTo automatically choose the best machine learning algorithm
   Reveal Solution Discussion Select an answer first
2. 7application · medium

   A data scientist is building a model to predict insurance claim amounts. Domain knowledge suggests that the effect of 'age' on claims is different for smokers versus non-smokers. The baseline model uses 'age' and 'is_smoker' as separate features. Which feature engineering step should be added to capture this combined effect?

   • ARemove 'is_smoker' and rely on 'age' alone to capture the pattern.
   • BCreate an interaction feature 'age * is_smoker' and add it to the model.
   • CBin 'age' into deciles and use the binned version instead of the raw value.
   • DStandardize 'age' and 'is_smoker' separately before training.
   Reveal Solution Discussion Select an answer first
4. 8foundation · easy

   Which of the following is a model-based method for assessing feature importance?

   • APermutation importance
   • BMutual information
   • CChi-square test
   • DCorrelation coefficient
   Reveal Solution Discussion Select an answer first
5. 9foundation · easy

   What is a potential risk of simply deleting rows that contain missing values?

   • AIt automatically encodes categorical variables
   • BIt increases the dimensionality of the dataset
   • CLoss of valuable information and potential introduction of bias
   • DIt guarantees the model will overfit
   Reveal Solution Discussion Select an answer first
6. 10application · medium

   A data scientist is forecasting daily store sales using a gradient boosting model. The raw data contains only the daily sales amount and the date. The scientist wants to capture weekly seasonality and recent trends. Which set of features should be created?

   • AA single feature equal to the number of days since the first record.
   • BDay-of-week, day-of-month, and a 7-day rolling average of sales.
   • CLag features at 1, 7, and 30 days without any calendar features.
   • DA 30-day rolling average and a 30-day rolling standard deviation only.
   Reveal Solution Discussion Select an answer first