4.2 Describe Automated Data Preparation Tasks (quality Checks, Formatting, Transformation, Cleaning) AI-TECHNICAL-PRACTITIONER Practice Questions (Page 3)

Questions 11–15

1. 11expert · hard

   A data team is automating the quality checks for a dataset that will be used to train a fraud detection model. The dataset is highly imbalanced, with very few fraudulent transactions. They need to ensure that the quality checks do not inadvertently remove the rare fraudulent cases. Which automated quality check strategy should they implement?

   • AReplace outliers with the mean value.
   • BFlag outliers using a robust method (e.g., IQR) and review flagged records manually before removal.
   • COnly check for missing values and duplicates; outliers are not a concern.
   • DAutomatically remove all records flagged as outliers.
   Reveal Solution Discussion Select an answer first
2. 12foundation · easy

   An automated pipeline removes rows that contain invalid or irrelevant data before analysis. Which data preparation task is being performed?

   • ACleaning
   • BQuality check
   • CTransformation
   • DFormatting
   Reveal Solution Discussion Select an answer first
4. 13foundation · easy

   Which transformation technique scales numeric data to a common range, such as 0 to 1, to prevent features with larger magnitudes from dominating analysis?

   • AEncoding
   • BAggregation
   • CFormatting
   • DNormalization
   Reveal Solution Discussion Select an answer first
5. 14application · medium

   A data engineer is automating the cleaning of a product inventory dataset. The dataset contains some records with negative quantities (which are invalid) and some records with duplicate product IDs. Which automated cleaning tasks should be applied?

   • AReplace negative quantities with zero and keep duplicates.
   • BRemove all records with negative quantities and keep duplicates for analysis.
   • COnly remove duplicate product IDs; negative quantities are acceptable.
   • DRemove records with negative quantities and remove duplicate product IDs.
   Reveal Solution Discussion Select an answer first
6. 15application · medium

   A data analyst is automating the quality checks for a survey dataset. The dataset includes a 'satisfaction_score' field (1-5) and a 'comments' field. They want to flag records where the satisfaction score is outside the valid range and where comments are missing. Which automated quality checks should they implement?

   • ACheck that satisfaction scores are between 1 and 5, and flag records with missing comments.
   • BOnly check for missing comments; satisfaction scores are always valid.
   • CReplace invalid scores with the average and fill missing comments with 'N/A'.
   • DRemove all records with missing comments and ignore invalid scores.
   Reveal Solution Discussion Select an answer first