Overfitting CT-AI Practice Questions (Page 2)

Questions 6–10

1. 6foundation · easy

   In the context of machine learning, what does the term 'overfitting' refer to?

   • AThe model performs poorly on both the training data and the test data.
   • BThe model performs well on the training data but fails to generalize to new, unseen data.
   • CThe model performs equally well on training and test data, indicating good generalization.
   • DThe model has too few parameters to capture the complexity of the data.
   Reveal Solution Discussion Select an answer first
2. 7application · medium

   A team is developing a model to predict loan defaults. They have a small dataset and are concerned about overfitting. They decide to use k-fold cross-validation to evaluate the model. What is the primary benefit of this approach?

   • AIt eliminates the need for a separate test set.
   • BIt automatically selects the best model architecture.
   • CIt increases the amount of training data by generating synthetic samples.
   • DIt provides a more reliable estimate of model performance on unseen data by using multiple train/test splits.
   Reveal Solution Discussion Select an answer first
4. 8application · medium

   A team is training a gradient-boosted tree model and notices that after 500 boosting rounds, training accuracy is 100% but validation accuracy has plateaued and started to decline. Which of the following actions is most appropriate to improve validation performance?

   • AIncrease the number of boosting rounds to allow the model to learn more patterns.
   • BRemove the validation set and train on all available data to improve generalization.
   • CIncrease the learning rate to help the model converge faster.
   • DApply early stopping to halt training when validation performance stops improving.
   Reveal Solution Discussion Select an answer first
5. 9foundation · easy

   Which scenario is most likely to lead to overfitting?

   • ATraining a model with regularization on a large dataset.
   • BTraining a high-capacity model on a small dataset.
   • CTraining a model with early stopping on a moderate dataset.
   • DTraining a simple model on a large dataset.
   Reveal Solution Discussion Select an answer first
6. 10application · medium

   A data scientist is evaluating a model and observes that the training error is very low, but the test error is high. According to the bias-variance tradeoff, this model is likely suffering from which of the following?

   • AHigh bias
   • BLow variance
   • CLow bias and low variance
   • DHigh variance
   Reveal Solution Discussion Select an answer first