AI Life Cycle (Model Selection, Training, and Validation) AAISM Practice Questions (Page 4)

Questions 16–20

1. 16foundation · easy

   In a medical diagnosis model, which metric is most important to prioritize when the cost of a false negative (missing a disease) is very high?

   • APrecision
   • BRecall
   • CF1-score
   • DAccuracy
   Reveal Solution Discussion Select an answer first
2. 17foundation · easy

   Which validation technique is most appropriate when the dataset is small and you want to maximize the use of data for both training and validation?

   • AK-fold cross-validation
   • BLeave-one-out validation
   • CTrain on the entire dataset and evaluate on the same data
   • DHoldout validation with a 70/30 split
   Reveal Solution Discussion Select an answer first
4. 18foundation · easy

   Which practice supports transparency and accountability in model validation?

   • AUsing only automated validation without human review
   • BRecording validation results and decisions, including model limitations
   • CDeleting validation data after the model is deployed
   • DKeeping validation results confidential to avoid scrutiny
   Reveal Solution Discussion Select an answer first
5. 19application · medium

   A team is training a linear regression model to predict housing prices. The model performs poorly on both training and test data, with high error on both. The team suspects the model is too simple to capture the underlying patterns. Which strategy would best address this issue?

   • AApply stronger regularization to the current model.
   • BIncrease the learning rate of the optimization algorithm.
   • CReduce the number of training samples to make the model easier to fit.
   • DIncrease model complexity by adding polynomial features or using a more flexible algorithm.
   Reveal Solution Discussion Select an answer first
6. 20application · medium

   A data science team is training a deep neural network for image classification. After several epochs, the training loss decreases steadily, but the validation loss begins to increase. The team needs to improve the model's generalization. Which action should they take first?

   • AReduce the size of the training dataset to simplify the model.
   • BAdd regularization, such as dropout or weight decay, to reduce overfitting.
   • CIncrease the number of training epochs to allow the model to converge further.
   • DSwitch to a more complex model architecture to capture more patterns.
   Reveal Solution Discussion Select an answer first