Oversee Data Evaluation PMI-CPMAI Practice Questions (Page 2)

Questions 6–10

1. 6expert · hard

   A data team is preparing a dataset for a natural language processing (NLP) model that classifies customer reviews as positive, negative, or neutral. The dataset contains a 'review_text' column and a 'rating' column (1-5 stars). During schema review, the team notices that the 'review_text' column contains HTML tags and special characters. What is the most appropriate action?

   • AConvert the HTML tags to plain text by rendering them.
   • BKeep the HTML tags because they may contain useful information.
   • CClean the text data by removing HTML tags and special characters during preprocessing.
   • DDrop the 'review_text' column and use only the 'rating' column.
   Reveal Solution Discussion Select an answer first
2. 7foundation · easy

   Which concept refers to how current or up-to-date the data is for the purpose of training an AI model?

   • AData consistency
   • BData relevance
   • CData completeness
   • DData freshness
   Reveal Solution Discussion Select an answer first
4. 8application · medium

   A data team is evaluating a dataset for a credit scoring model. The dataset contains a 'credit_score' column, but during EDA they find that 20% of the values are negative, which is impossible for the credit score range (300-850). What should the team do first?

   • AReplace all negative values with the mean credit score.
   • BRemove all records with negative credit scores.
   • CConvert negative values to positive by taking the absolute value.
   • DInvestigate the source of the negative values to determine if they are data entry errors or a different coding scheme.
   Reveal Solution Discussion Select an answer first
5. 9expert · hard

   A data governance team is evaluating a dataset for a model that predicts patient readmission. The dataset is assembled from multiple hospitals, and each hospital has its own data entry standards. The team finds that the 'diagnosis_code' column uses different coding systems: some hospitals use ICD-9 and others use ICD-10. The model requires a single consistent coding system. What is the most appropriate action?

   • ADrop the diagnosis_code column because it is inconsistent.
   • BMap all ICD-9 codes to ICD-10 using a validated mapping table.
   • CKeep both coding systems and treat them as separate features.
   • DUse the ICD-9 codes only because they are more common.
   Reveal Solution Discussion Select an answer first
6. 10application · medium

   A logistics company is building a model to predict delivery delays. The available data includes weather conditions, traffic data, and historical delivery times. The weather data is updated hourly, but the traffic data is only updated daily. The model is expected to make predictions in near real-time. What is the most significant data evaluation concern?

   • AThe historical delivery times are not accurate.
   • BThe traffic data may not be fresh enough to capture current conditions, reducing the model's relevance for real-time predictions.
   • CThe schema of the traffic data is incompatible with the weather data.
   • DThe weather data is too fresh and may introduce noise.
   Reveal Solution Discussion Select an answer first