Data Science FREE Data Science Model Evaluation and Validation Questions and Answers 2

Question 1

Which metric is most appropriate for evaluating a classification model on a highly imbalanced dataset?

Accepted Answer

Precision-Recall AUC

Answer

Accuracy

Answer

Mean Squared Error

Answer

R-squared

Question 2

What does a high variance and low bias in a model typically indicate?

Accepted Answer

Overfitting

Answer

Underfitting

Answer

Optimal generalization

Answer

Insufficient data preprocessing

Question 3

In stratified k-fold cross-validation, what is preserved across each fold?

Accepted Answer

The proportion of each class label

Answer

The number of features

Answer

The order of data points

Answer

The ratio of training to test size

Question 4

What is the primary purpose of a calibration curve (reliability diagram)?

Accepted Answer

To assess whether predicted probabilities match actual outcomes

Answer

To measure feature importance

Answer

To detect multicollinearity

Answer

To optimize hyperparameters

Question 5

When using bootstrapping for model evaluation, what is the typical approach?

Accepted Answer

Sampling with replacement to create multiple training sets

Answer

Splitting data into exactly two halves

Answer

Removing outliers before each evaluation

Answer

Using only the first 80% of data chronologically

Question 6

What does the Kolmogorov-Smirnov (KS) statistic measure in model evaluation?

Accepted Answer

The maximum separation between cumulative distributions of positive and negative classes

Answer

The correlation between predicted and actual values

Answer

The average log-likelihood of predictions

Answer

The number of misclassified samples