Objective 3.4 Evaluate the Model AIP-210 Practice Questions (Page 6)

Questions 26–30

1. 26application · medium

   A data scientist has trained two models, A and B, on the same dataset. Using 5-fold cross-validation, the mean accuracy of Model A is 0.85 with a standard deviation of 0.02, and Model B is 0.84 with a standard deviation of 0.05. Which conclusion is most appropriate?

   • AModel A is definitively better because it has higher mean accuracy
   • BModel A is more stable, but statistical testing is needed to determine if the accuracy difference is significant
   • CModel B is better because it has higher variance, indicating it explores more of the hypothesis space
   • DThe models are equivalent because the mean accuracies are close
   Reveal Solution Discussion Select an answer first
2. 27application · medium

   A data scientist is evaluating a small dataset of 500 samples for a binary classification problem. The dataset is imbalanced, with 90% negative class. The scientist wants to assess model stability and compare two algorithms. Which cross-validation approach is most appropriate?

   • AStratified k-fold cross-validation with k=5
   • BSimple k-fold cross-validation with k=10
   • CLeave-one-out cross-validation
   • DHoldout validation with a 70/30 split
   Reveal Solution Discussion Select an answer first
4. 28foundation · easy

   What is the primary risk of using the test set multiple times during model development?

   • AThe model may overfit to the test set, leading to overly optimistic performance estimates
   • BThe training set will be contaminated
   • CThe test set will be deleted
   • DThe test set will become too large
   Reveal Solution Discussion Select an answer first
5. 29foundation · easy

   For a clustering model, which metric evaluates the quality of clusters by measuring the average silhouette width of all samples?

   • ACalinski-Harabasz index
   • BSilhouette score
   • CAdjusted Rand Index
   • DDavies-Bouldin index
   Reveal Solution Discussion Select an answer first
6. 30foundation · easy

   Which statement best describes the bias-variance tradeoff?

   • ABias and variance are independent and can both be minimized without affecting each other
   • BIncreasing model complexity always reduces bias and variance simultaneously
   • CThere is a tradeoff between error due to bias and error due to variance, and the goal is to find a balance that minimizes total error
   • DThe tradeoff only applies to classification models, not regression models
   Reveal Solution Discussion Select an answer first