AI Life Cycle (Model Selection, Training, and Validation) AAISM Practice Questions (Page 1)

Questions 1–5

1. 1foundation · easy

   What is a key governance principle for model selection in an organization?

   • ASelecting the model with the highest accuracy regardless of other factors
   • BChoosing the most complex model to ensure future-proofing
   • CKeeping model selection decisions informal to speed up development
   • DDocumenting the rationale for model selection and ensuring alignment with organizational policies
   Reveal Solution Discussion Select an answer first
2. 2foundation · easy

   A model performs very well on the training data but poorly on the validation data. What is this an indication of?

   • AData leakage
   • BOverfitting
   • CClass imbalance
   • DUnderfitting
   Reveal Solution Discussion Select an answer first
4. 3application · medium

   A data science team is building a model to predict loan defaults. They have historical data that includes a 'date' column. They plan to split the data randomly into training and test sets. The compliance officer is concerned about data leakage. Which action should the team take to ensure unbiased evaluation?

   • ASplit the data chronologically, using earlier data for training and later data for testing.
   • BUse cross-validation with random shuffling to average out temporal effects.
   • CUse a random split and ensure no duplicate records appear in both sets.
   • DRemove the date column entirely to avoid any leakage.
   Reveal Solution Discussion Select an answer first
5. 4foundation · easy

   What is data leakage in the context of data splitting?

   • AThe training data is accidentally deleted
   • BThe test set is too small to provide reliable results
   • CThe model is trained on too much data
   • DInformation from the test set is used during training, leading to overly optimistic performance estimates
   Reveal Solution Discussion Select an answer first
6. 5foundation · easy

   Which strategy is most effective for mitigating underfitting?

   • AReducing the number of features
   • BAdding more regularization
   • CDecreasing the training data size
   • DIncreasing model complexity
   Reveal Solution Discussion Select an answer first