Evaluation and Validation MCQs

1. What is the purpose of cross-validation in machine learning?

A. To reduce the size of the dataset
B. To evaluate the model’s performance and reduce overfitting
C. To improve the model’s accuracy by adding more features
D. To increase the number of training examples

Answer: B


2. Which of the following is a common evaluation metric for classification models?

A. Root Mean Squared Error (RMSE)
B. F1 Score
C. Adjusted R-Squared
D. Mean Absolute Error (MAE)

Answer: B


3. In a confusion matrix, what does the term “True Positive” (TP) refer to?

A. The number of instances that were incorrectly classified as negative
B. The number of instances that were correctly classified as positive
C. The number of instances that were incorrectly classified as positive
D. The number of instances that were correctly classified as negative

Answer: B


4. Which of the following is used to assess the performance of a regression model?

A. Accuracy
B. Precision
C. Mean Squared Error (MSE)
D. F1 Score

Answer: C


5. Which of the following is a technique used to evaluate the performance of a model on imbalanced datasets?

A. Confusion Matrix
B. Accuracy
C. ROC Curve and AUC
D. Root Mean Squared Error (RMSE)

Answer: C


6. What does the “AUC” (Area Under the Curve) represent in the context of ROC (Receiver Operating Characteristic) analysis?

A. The number of positive predictions made by the model
B. The model’s ability to correctly classify both positive and negative instances
C. The threshold value for decision-making
D. The amount of data used in the training set

Answer: B


7. What is a common issue when using accuracy as a performance metric for classification models on imbalanced datasets?

A. Accuracy can be misleading and not reflect model performance for the minority class
B. Accuracy always overestimates the model’s performance
C. Accuracy cannot handle missing values
D. Accuracy is not suitable for multiclass classification

Answer: A


8. In k-fold cross-validation, what is the typical purpose of splitting the dataset into k subsets?

A. To test the model on a smaller dataset
B. To ensure the model is evaluated on all portions of the dataset and prevent overfitting
C. To speed up the model training process
D. To balance the dataset

Answer: B


9. What does the term “overfitting” refer to in model evaluation?

A. The model is too simple and underperforms on the training data
B. The model performs well on both the training and test data
C. The model performs excellently on training data but poorly on unseen data due to excessive complexity
D. The model performs well on the test data but not on the training data

Answer: C


10. What is the primary goal of model validation?

A. To adjust the parameters of the model
B. To improve the model’s performance on unseen data
C. To test the model’s speed
D. To reduce the size of the training data

Answer: B

Leave a Reply

Your email address will not be published. Required fields are marked *