NSCT – Model Evaluation & Validation MCQs 20 min Score: 0 Attempted: 0/20 Subscribe 1. . Model evaluation in machine learning is: (A) Encrypting models only (B) The process of assessing how well a trained model performs on unseen data (C) Compressing datasets only (D) Backup onlyShow All Answers 2. . Validation of a model is done to: (A) Tune hyperparameters and check performance before final testing (B) Encrypt validation data (C) Compress features (D) Backup only 3. . Training set is used to: (A) Fit the model by learning patterns from data (B) Encrypt data (C) Compress data (D) Backup only 4. . Test set is used to: (A) Compress test set (B) Encrypt test data (C) Evaluate the model's performance on unseen data (D) Backup only 5. . Cross-validation is: (A) A technique to assess model performance by dividing data into multiple training and validation sets (B) Encrypting cross data (C) Compressing validation sets (D) Backup only 6. . K-Fold cross-validation divides data into: (A) Encrypts K folds (B) Two parts only (C) K subsets where each subset is used as a validation set once (D) Backup only 7. . Leave-One-Out Cross-Validation (LOOCV) is: (A) Encrypting data points (B) Using one data point as validation and the rest for training, repeated for all points (C) Compressing data points (D) Backup only 8. . Accuracy is: (A) Backup only (B) Encrypting accuracy (C) Compressing accuracy (D) The ratio of correctly predicted instances to total instances 9. . Precision measures: (A) Backup only (B) Encrypting precision (C) Compressing precision (D) The proportion of true positives among all predicted positives 10. . Recall (Sensitivity) measures: (A) Backup only (B) Encrypting recall (C) Compressing recall (D) The proportion of true positives among all actual positives 11. . F1-Score is: (A) Encrypting F1 (B) The harmonic mean of precision and recall (C) Compressing F1 (D) Backup only 12. . Confusion matrix shows: (A) Backup only (B) Encrypting matrix (C) Compressing matrix (D) True positives, false positives, true negatives, and false negatives 13. . ROC Curve is used to: (A) Evaluate the trade-off between true positive rate and false positive rate (B) Encrypt ROC (C) Compress ROC (D) Backup only 14. . AUC (Area Under Curve) represents: (A) The model's ability to discriminate between classes (B) Encrypting AUC (C) Compressing AUC (D) Backup only 15. . Mean Squared Error (MSE) is used for: (A) Compressing errors (B) Encrypting errors (C) Evaluating regression models by measuring average squared difference between actual and predicted values (D) Backup only 16. . R-squared (Coefficient of Determination) measures: (A) Proportion of variance in the dependent variable explained by the model (B) Encrypting R-squared (C) Compressing R-squared (D) Backup only 17. . Overfitting occurs when: (A) Compressing model (B) Encrypting overfitting (C) A model performs well on training data but poorly on unseen data (D) Backup only 18. . Underfitting occurs when: (A) Encrypting underfitting (B) A model is too simple to capture underlying patterns (C) Compressing model (D) Backup only 19. . Early stopping in model validation helps to: (A) Backup only (B) Encrypt training (C) Compress iterations (D) Prevent overfitting by stopping training when performance on validation set stops improving 20. . The main purpose of model evaluation and validation is to: (A) Ensure the model generalizes well, is accurate, and reliable before deployment (B) Encrypt models (C) Compress data (D) Backup only
