Objective 3.3 Train, Validate, and Test Data Subsets AIP-210 Practice Questions (Page 1)

Questions 1–5

1. 1application · medium

   A junior data scientist is setting up an experiment for a classification task. The scientist splits the data into training and test sets, then repeatedly evaluates different model configurations on the test set to choose the best one. What is the main problem with this approach?

   • AThe model is overfitting the training data
   • BThe test set is being used for model selection, which biases the final performance estimate
   • CThe data should be split into more than two subsets
   • DThe training set is too small to fit the model properly
   Reveal Solution Discussion Select an answer first
2. 2foundation · easy

   Which practice helps prevent data leakage during data splitting?

   • AUsing the entire dataset to compute normalization statistics before splitting
   • BShuffling the data before splitting to ensure randomness
   • CFitting scalers and imputers on the training subset only, then applying them to validation and test subsets
   • DSplitting the data after performing feature selection on the full dataset
   Reveal Solution Discussion Select an answer first
4. 3foundation · easy

   What is the primary purpose of partitioning a dataset into training, validation, and test subsets?

   • ATo ensure the model trains faster by using smaller batches
   • BTo reduce the total amount of data needed for model training
   • CTo guarantee the model achieves high accuracy on the training data
   • DTo enable unbiased evaluation of model performance on unseen data
   Reveal Solution Discussion Select an answer first
5. 4foundation · easy

   When should the test subset be used in the model development process?

   • ADuring each training epoch to monitor loss
   • BDuring hyperparameter tuning to compare models
   • CAfter hyperparameter tuning and model selection are complete
   • DBefore training begins to set initial parameters
   Reveal Solution Discussion Select an answer first
6. 5foundation · easy

   What is the primary role of the validation subset in model development?

   • ATo provide a final unbiased estimate of generalization performance
   • BTo tune hyperparameters and select the best model
   • CTo fit the model's parameters during training
   • DTo increase the size of the training set
   Reveal Solution Discussion Select an answer first