Developing End-To-End ML Pipelines PROFESSIONAL-MACHINE-LEARNING-ENGINEER Practice Questions (Page 6)

Questions 26–28

1. 26application · medium

   A data science team is training a model to predict whether a transaction is fraudulent. The dataset is highly imbalanced (1% fraud). They have split the data into training, validation, and holdout sets. They want to select the best model based on its ability to identify fraud cases. Which metric should they use for model selection?

   • AAUC on the training set
   • BAccuracy on the validation set
   • CLog loss on the holdout set
   • DF1-score on the validation set
   Reveal Solution Discussion Select an answer first
2. 27foundation · easy

   Which open-source tool is commonly used for building custom, unmanaged ML pipeline orchestration on Google Cloud?

   • ARay
   • BCloud Composer
   • CBigQuery
   • DVertex AI Pipelines
   Reveal Solution Discussion Select an answer first
3. 28foundation · easy

   Which of the following is a typical check performed during data validation in an ML pipeline?

   • AVerifying that the data schema matches the expected feature names and types
   • BTuning hyperparameters to improve performance
   • CDeploying the model to a production endpoint
   • DCalculating the model's accuracy on a test set
   Reveal Solution Discussion Select an answer first