Correct Answer: A negative case predicted as positive
Explanation:
A false positive occurs when the model predicts the positive class for an example that is actually negative.
Incorrect! Try again.
13What is the main purpose of cross-validation?
Model Improvement (Cross Validation, Hyperparameter Tuning, Grid Search, Random Search)
Easy
A.To remove every feature from the dataset
B.To increase the number of target labels
C.To guarantee perfect predictions
D.To estimate how well a model generalizes
Correct Answer: To estimate how well a model generalizes
Explanation:
Cross-validation evaluates a model on different portions of the data to estimate performance on unseen examples.
Incorrect! Try again.
14In -fold cross-validation, what happens to the dataset?
Model Improvement (Cross Validation, Hyperparameter Tuning, Grid Search, Random Search)
Easy
A.It is reduced to one training example
B.It is converted into target variables
C.It is divided into parts
D.It is sorted only by prediction value
Correct Answer: It is divided into parts
Explanation:
In -fold cross-validation, the data is split into folds, which are used for training and validation in rotation.
Incorrect! Try again.
15What is a hyperparameter?
Model Improvement (Cross Validation, Hyperparameter Tuning, Grid Search, Random Search)
Easy
A.An error calculated from the test set
B.A prediction produced after model training
C.A label created by the confusion matrix
D.A setting chosen before model training
Correct Answer: A setting chosen before model training
Explanation:
Hyperparameters are model settings, such as tree depth or learning rate, selected before training begins.
Incorrect! Try again.
16What does hyperparameter tuning try to find?
Model Improvement (Cross Validation, Hyperparameter Tuning, Grid Search, Random Search)
Easy
A.The exact value of every missing feature
B.A new classification label for each sample
C.The names of all training examples
D.A useful combination of model settings
Correct Answer: A useful combination of model settings
Explanation:
Hyperparameter tuning tests different settings to identify a combination that gives strong validation performance.
Incorrect! Try again.
17How does grid search usually select hyperparameter values?
Model Improvement (Cross Validation, Hyperparameter Tuning, Grid Search, Random Search)
Easy
A.It changes the target labels during training
B.It chooses values without evaluating the model
C.It tests only the first combination
D.It tests all specified combinations
Correct Answer: It tests all specified combinations
Explanation:
Grid search evaluates every combination in the predefined hyperparameter grid.
Incorrect! Try again.
18How does random search choose hyperparameter combinations?
Model Improvement (Cross Validation, Hyperparameter Tuning, Grid Search, Random Search)
Easy
A.By sampling combinations from specified ranges
B.By using the test labels to set each parameter
C.By testing every possible combination in order
D.By selecting only the smallest parameter values
Correct Answer: By sampling combinations from specified ranges
Explanation:
Random search evaluates randomly selected combinations from the given hyperparameter distributions or ranges.
Incorrect! Try again.
19What is the main goal of Explainable AI (XAI)?
Introduction to Explainable AI
Easy
A.To make model decisions easier for people to understand
B.To ensure that every model has zero prediction errors
C.To replace all datasets with manually written rules
D.To make models use only numerical input features
Correct Answer: To make model decisions easier for people to understand
Explanation:
Explainable AI provides understandable reasons or information about how a model reaches its predictions.
Incorrect! Try again.
20Why can explainability be important in an AI system?
Introduction to Explainable AI
Easy
A.It removes the need to evaluate model performance
B.It can help users identify and trust appropriate model behavior
C.It guarantees that the training data contains no bias
D.It makes every prediction independent of input data
Correct Answer: It can help users identify and trust appropriate model behavior
Explanation:
Explanations can help people understand, check, and responsibly use a model's predictions.
Incorrect! Try again.
21A regression model produces errors of , , , and . What is the model's MAE?
Regression Metrics (MAE, MSE, RMSE, R² Score)
Medium
A.
B.
C.
D.
Correct Answer:
Explanation:
MAE is the mean of the absolute errors: .
Incorrect! Try again.
22Two regression models have error sets. Model A has errors , while Model B has errors . Which model has the lower MAE?
Regression Metrics (MAE, MSE, RMSE, R² Score)
Medium
A.Model A, because its MAE is
B.Model A, because its MAE is
C.Model B, because its MAE is
D.Both models, because their MAEs are equal
Correct Answer: Both models, because their MAEs are equal
Explanation:
Model A has MAE , and Model B has MAE . Therefore, their MAEs are equal.
Incorrect! Try again.
23A model's residuals are . What is its MSE?
Regression Metrics (MAE, MSE, RMSE, R² Score)
Medium
A.
B.
C.
D.
Correct Answer:
Explanation:
MSE is the average of squared errors: .
Incorrect! Try again.
24A regression model has an MSE of . What is its RMSE, and what does this value represent?
Regression Metrics (MAE, MSE, RMSE, R² Score)
Medium
A., the typical error in the target's units
B., the average absolute error
C., the typical error in squared units
D., the total squared error
Correct Answer: , the typical error in the target's units
Explanation:
RMSE is the square root of MSE, so . It is expressed in the same units as the target variable.
Incorrect! Try again.
25A regression model has . Which interpretation is most appropriate?
Regression Metrics (MAE, MSE, RMSE, R² Score)
Medium
A.The model has an average error of units
B.The model correctly predicts of observations
C.The model makes predictions with accuracy
D.The model explains of the target variance
Correct Answer: The model explains of the target variance
Explanation:
measures the proportion of variance in the target explained by the model relative to a baseline mean prediction.
Incorrect! Try again.
26A classifier produces true positives, false positives, false negatives, and true negatives. What is its precision?
Confusion Matrix and Classification Metrics (Accuracy, Precision, Recall, F1-score, ROC-AUC)
Medium
A.
B.
C.
D.
Correct Answer:
Explanation:
Precision is .
Incorrect! Try again.
27Using true positives, false positives, false negatives, and true negatives, what is the classifier's recall?
Confusion Matrix and Classification Metrics (Accuracy, Precision, Recall, F1-score, ROC-AUC)
Medium
A.
B.
C.
D.
Correct Answer:
Explanation:
Recall is .
Incorrect! Try again.
28A medical screening model has high recall but low precision. What does this most likely mean?
Confusion Matrix and Classification Metrics (Accuracy, Precision, Recall, F1-score, ROC-AUC)
Medium
A.It misses most positive cases
B.It identifies most positives but raises many false alarms
C.It classifies both classes with equal accuracy
D.It produces few positive predictions overall
Correct Answer: It identifies most positives but raises many false alarms
Explanation:
High recall means few actual positives are missed, while low precision means many predicted positives are false positives.
Incorrect! Try again.
29A classifier has precision and recall . What is its F1-score?
Confusion Matrix and Classification Metrics (Accuracy, Precision, Recall, F1-score, ROC-AUC)
Medium
A.
B.
C.
D.
Correct Answer:
Explanation:
The F1-score is the harmonic mean: .
Incorrect! Try again.
30A fraud dataset contains fraudulent transactions. A model predicts every transaction as legitimate and achieves accuracy. Which metric would better reveal its failure to detect fraud?
Confusion Matrix and Classification Metrics (Accuracy, Precision, Recall, F1-score, ROC-AUC)
Medium
A.Recall for the fraud class
B.The number of true negatives
C.The total number of observations
D.Accuracy
Correct Answer: Recall for the fraud class
Explanation:
The model has no true positives, so its fraud recall is . Recall exposes the failure that accuracy hides because of class imbalance.
Incorrect! Try again.
31What does an ROC-AUC value of generally indicate for a binary classifier?
Confusion Matrix and Classification Metrics (Accuracy, Precision, Recall, F1-score, ROC-AUC)
Medium
A.Correct predictions for half the samples
B.Performance similar to random ranking
C.Perfect classification
D.A model with no false positives
Correct Answer: Performance similar to random ranking
Explanation:
An AUC of means the model generally ranks positive and negative examples no better than random chance.
Incorrect! Try again.
32When the cost of missing a positive case is much higher than the cost of a false alarm, which metric should receive greater emphasis?
Confusion Matrix and Classification Metrics (Accuracy, Precision, Recall, F1-score, ROC-AUC)
Medium
A.Specificity
B.Recall
C.Overall accuracy
D.Precision
Correct Answer: Recall
Explanation:
Recall measures the proportion of actual positive cases detected, so it is especially important when false negatives are costly.
Incorrect! Try again.
33In 5-fold cross-validation, how is the dataset typically used during one complete evaluation cycle?
Model Improvement (Cross Validation, Hyperparameter Tuning, Grid Search, Random Search)
Medium
A.Each fold is used only for final testing
B.It is split into five parts, each used once for validation
C.Five separate datasets are collected for testing
D.The model is trained only on one part
Correct Answer: It is split into five parts, each used once for validation
Explanation:
In 5-fold cross-validation, the model trains on four folds and validates on the remaining fold, repeating this until every fold has served as validation data.
Incorrect! Try again.
34Why is cross-validation often preferred to a single train-validation split when the dataset is small?
Model Improvement (Cross Validation, Hyperparameter Tuning, Grid Search, Random Search)
Medium
A.It uses multiple validation partitions for a more stable estimate
B.It always increases the final training accuracy
C.It guarantees that the model cannot overfit
D.It removes the need for a test set
Correct Answer: It uses multiple validation partitions for a more stable estimate
Explanation:
Cross-validation averages performance across several splits, reducing dependence on one potentially unrepresentative validation partition.
Incorrect! Try again.
35Which of the following is a hyperparameter for a decision tree model?
Model Improvement (Cross Validation, Hyperparameter Tuning, Grid Search, Random Search)
Medium
A.The predicted class label
B.The residual for one observation
C.The tree's maximum depth
D.A training example's feature value
Correct Answer: The tree's maximum depth
Explanation:
Maximum depth is selected before training and controls model complexity, so it is a hyperparameter.
Incorrect! Try again.
36A grid search tests 3 values for a learning rate and 4 values for maximum tree depth. Ignoring cross-validation repetitions, how many hyperparameter combinations are evaluated?
Model Improvement (Cross Validation, Hyperparameter Tuning, Grid Search, Random Search)
Medium
A.
B.
C.
D.
Correct Answer:
Explanation:
Grid search evaluates every combination, giving combinations.
Incorrect! Try again.
37Why can random search be more efficient than grid search when only a few hyperparameters strongly affect performance?
Model Improvement (Cross Validation, Hyperparameter Tuning, Grid Search, Random Search)
Medium
A.It eliminates the need for validation data
B.It guarantees the global optimum
C.It tests every possible combination
D.It samples combinations across the search space
Correct Answer: It samples combinations across the search space
Explanation:
Random search can explore more useful values of influential hyperparameters without spending equal effort on every grid dimension.
Incorrect! Try again.
38A model performs very well on training folds but poorly on validation folds during cross-validation. Which change is most likely to help?
Model Improvement (Cross Validation, Hyperparameter Tuning, Grid Search, Random Search)
Medium
A.Train and evaluate on the same samples
B.Remove the validation folds
C.Use regularization or reduce model complexity
D.Increase model complexity substantially
Correct Answer: Use regularization or reduce model complexity
Explanation:
This pattern suggests overfitting. Regularization or a simpler model can improve generalization to unseen data.
Incorrect! Try again.
39What is the primary goal of Explainable AI?
Introduction to Explainable AI
Medium
A.To make model decisions understandable to people
B.To make every model smaller
C.To guarantee predictions are unbiased
D.To replace evaluation metrics
Correct Answer: To make model decisions understandable to people
Explanation:
Explainable AI provides understandable reasons or evidence for model outputs, supporting interpretation, trust, and auditing.
Incorrect! Try again.
40A feature-importance analysis shows that income contributes strongly to a loan model's predictions. What does this result mean?
Introduction to Explainable AI
Medium
A.Income alone determines every prediction
B.Income causes the approved loans
C.Income is associated with the model's predictive behavior
D.Income must be the fairest feature
Correct Answer: Income is associated with the model's predictive behavior
Explanation:
Feature importance describes how much a feature contributes to predictions in the model. It does not prove causation or guarantee fairness.
Incorrect! Try again.
41A regression model has residuals . What are its MAE, MSE, and RMSE, respectively?
Regression Metrics (MAE, MSE, RMSE, R² Score)
Hard
A.
B.
C.
D.
Correct Answer:
Explanation:
MAE is the mean absolute residual: . MSE is , and RMSE is .
Incorrect! Try again.
42Two models are evaluated on the same regression targets. Model A has MAE and RMSE , while Model B has MAE and RMSE . Which conclusion is most defensible?
Regression Metrics (MAE, MSE, RMSE, R² Score)
Hard
A.Model A has fewer large errors than Model B
B.The metrics cannot be compared because their units differ
C.Model B is better under absolute-error loss
D.Model B has uniformly smaller prediction errors
Correct Answer: Model B is better under absolute-error loss
Explanation:
Lower MAE means lower average absolute error, so Model B is better under MAE. Its larger RMSE indicates that it likely has more severe outliers or larger individual errors.
Incorrect! Try again.
43A model obtains on a test set. The target variance is later found to be much smaller in a second test set, while the model's RMSE remains unchanged. What is the most likely effect on ?
Regression Metrics (MAE, MSE, RMSE, R² Score)
Hard
A.It remains exactly
B.It decreases because the residual variance is larger relative to target variance
C.It becomes zero because the target variance is smaller
D.It increases because RMSE is unchanged
Correct Answer: It decreases because the residual variance is larger relative to target variance
Explanation:
. With similar prediction error but smaller target variance, decreases, making the error ratio larger and reducing .
Incorrect! Try again.
44A test set contains one extreme target value caused by a measurement error. Which metric is most likely to change disproportionately if the model's prediction for that point is poor?
Regression Metrics (MAE, MSE, RMSE, R² Score)
Hard
A.R², because it ignores residual magnitude
B.MAE, because it uses absolute deviations
C.The median absolute error, because it emphasizes extremes
D.MSE, because it squares deviations
Correct Answer: MSE, because it squares deviations
Explanation:
Squaring residuals gives very large errors disproportionately high influence. RMSE and can also be affected, but MSE directly applies the squared penalty.
Incorrect! Try again.
45A regression model predicts the mean target value for every test example. Under ordinary definitions, which statement is correct when the test target variance is nonzero?
Regression Metrics (MAE, MSE, RMSE, R² Score)
Hard
A.Its is and its RMSE is zero
B.Its is undefined because constant predictions are invalid
C.Its is and its RMSE equals the target standard deviation
D.Its is negative and its MAE must equal zero
Correct Answer: Its is and its RMSE equals the target standard deviation
Explanation:
Predicting the test-set mean gives SSE equal to SST, so . The resulting MSE is the target variance, making RMSE equal to the target standard deviation.
Incorrect! Try again.
46A classifier produces , , , and . Which pair is correct?
Confusion Matrix and Classification Metrics (Accuracy, Precision, Recall, F1-score, ROC-AUC)
Hard
A.Precision , recall
B.Precision , recall
C.Precision , recall
D.Precision , recall
Correct Answer: Precision , recall
Explanation:
Precision is . Recall is .
Incorrect! Try again.
47In a highly imbalanced fraud-detection problem, a classifier predicts every transaction as legitimate. Which metric can appear excellent while the model remains operationally useless?
Confusion Matrix and Classification Metrics (Accuracy, Precision, Recall, F1-score, ROC-AUC)
Hard
A.Fraud F1-score
B.Fraud recall
C.Fraud precision
D.Accuracy
Correct Answer: Accuracy
Explanation:
If legitimate transactions dominate the data, always predicting the majority class yields high accuracy while producing zero recall for fraud.
Incorrect! Try again.
48For a binary classifier, lowering the decision threshold generally produces which change, assuming the score ordering stays fixed?
Confusion Matrix and Classification Metrics (Accuracy, Precision, Recall, F1-score, ROC-AUC)
Hard
A.Fewer predicted positives and higher precision
B.No change in the confusion matrix or threshold metrics
C.More predicted negatives and higher recall
D.More predicted positives and higher recall
Correct Answer: More predicted positives and higher recall
Explanation:
A lower threshold labels more examples as positive. This usually increases true positives and recall, while also potentially increasing false positives and reducing precision.
Incorrect! Try again.
49A classifier has precision and recall . What is its F1-score?
Confusion Matrix and Classification Metrics (Accuracy, Precision, Recall, F1-score, ROC-AUC)
Hard
A.
B.
C.
D.
Correct Answer:
Explanation:
.
Incorrect! Try again.
50A model's ROC-AUC is , but its precision at the deployment threshold is only $0.12. Which explanation is most plausible?
Confusion Matrix and Classification Metrics (Accuracy, Precision, Recall, F1-score, ROC-AUC)
Hard
A.The model ranks examples well, but positives are rare and the threshold is poorly selected
B.ROC-AUC directly equals precision averaged over thresholds
C.The model must have an invalid ROC curve
D.High ROC-AUC guarantees high precision at every threshold
Correct Answer: The model ranks examples well, but positives are rare and the threshold is poorly selected
Explanation:
ROC-AUC measures ranking across thresholds. Precision depends strongly on class prevalence and the selected operating threshold, so excellent AUC does not guarantee high precision.
Incorrect! Try again.
51Two classifiers have identical ROC-AUC values. Classifier A has higher precision at the threshold required by the application. What can be concluded?
Confusion Matrix and Classification Metrics (Accuracy, Precision, Recall, F1-score, ROC-AUC)
Hard
A.Classifier A is preferable for that operating requirement
B.Classifier A must have higher accuracy on every dataset
C.Classifier B has a strictly better ranking function
D.Classifier B has higher recall at every threshold
Correct Answer: Classifier A is preferable for that operating requirement
Explanation:
Equal ROC-AUC does not determine performance at one specific threshold. If precision is the relevant requirement, the classifier with higher precision at the deployed threshold is preferable.
Incorrect! Try again.
52A screening system must minimize missed positive cases, and false positives can be investigated manually. Which evaluation priority is most appropriate?
Confusion Matrix and Classification Metrics (Accuracy, Precision, Recall, F1-score, ROC-AUC)
Hard
A.Maximize precision, then ignore recall
B.Maximize ROC-AUC at one fixed threshold
C.Maximize accuracy because costs are symmetric
D.Maximize recall, then monitor precision
Correct Answer: Maximize recall, then monitor precision
Explanation:
When false negatives are especially costly, recall should be prioritized. Precision must still be monitored because excessive false positives can overwhelm the investigation process.
Incorrect! Try again.
53A multiclass classifier has per-class recalls of , , and . Which averaging strategy gives each class equal importance regardless of its sample count?
Confusion Matrix and Classification Metrics (Accuracy, Precision, Recall, F1-score, ROC-AUC)
Hard
A.Micro-averaged recall
B.Macro-averaged recall
C.Majority-class recall
D.Weighted-average recall
Correct Answer: Macro-averaged recall
Explanation:
Macro averaging computes the metric independently for each class and then gives every class equal weight, making it sensitive to minority-class performance.
Incorrect! Try again.
54A classifier's accuracy rises after deployment because the positive class becomes much rarer, while its positive recall falls sharply. Which interpretation is strongest?
Confusion Matrix and Classification Metrics (Accuracy, Precision, Recall, F1-score, ROC-AUC)
Hard
A.Recall must be wrong because accuracy and recall always move together
B.The classifier became perfectly calibrated
C.The class distribution changed, making accuracy misleading for the positive class
D.The model necessarily improved because accuracy increased
Correct Answer: The class distribution changed, making accuracy misleading for the positive class
Explanation:
Accuracy is sensitive to class prevalence. A rarer positive class can increase accuracy even as the model misses a larger fraction of actual positives, so class-specific metrics are necessary.
Incorrect! Try again.
55Why should preprocessing steps such as feature scaling be fitted separately inside each training fold during cross-validation?
Model Improvement (Cross Validation, Hyperparameter Tuning, Grid Search, Random Search)
Hard
A.To guarantee that all models use identical coefficients
B.To make every fold contain the same number of samples
C.To prevent validation-fold information from influencing the fitted transformation
D.To increase the validation score through additional data leakage
Correct Answer: To prevent validation-fold information from influencing the fitted transformation
Explanation:
Fitting preprocessing on the full dataset allows validation statistics to leak into training. A pipeline fits transformations only on each fold's training portion and then applies them to that fold's validation portion.
Incorrect! Try again.
56A dataset contains repeated measurements from the same patients. Randomly assigning rows to cross-validation folds causes very high validation scores. Which change best addresses the problem?
Model Improvement (Cross Validation, Hyperparameter Tuning, Grid Search, Random Search)
Hard
A.Use more random folds with the same row-level split
B.Use the test set during hyperparameter selection
C.Use group-based cross-validation with patient identity as the group
D.Remove regularization so patient patterns are retained
Correct Answer: Use group-based cross-validation with patient identity as the group
Explanation:
Rows from one patient can share highly specific information. Group-based splitting keeps all measurements from a patient in one fold, producing a more realistic estimate of generalization to unseen patients.
Incorrect! Try again.
57A grid search evaluates five values for each of four independent hyperparameters. How many configurations are evaluated before accounting for cross-validation folds?
Model Improvement (Cross Validation, Hyperparameter Tuning, Grid Search, Random Search)
Hard
A.20
B.125
C.625
D.25
Correct Answer: 625
Explanation:
Grid search evaluates every Cartesian-product combination: configurations.
Incorrect! Try again.
58When is random search often more efficient than grid search for hyperparameter optimization?
Model Improvement (Cross Validation, Hyperparameter Tuning, Grid Search, Random Search)
Hard
A.When all hyperparameters affect performance equally
B.When every possible configuration must be tested exhaustively
C.When only a few hyperparameters strongly influence performance
D.When the search space contains no continuous values
Correct Answer: When only a few hyperparameters strongly influence performance
Explanation:
Grid search spends many trials varying weakly influential dimensions. Random search explores more distinct values in the important dimensions and can find strong configurations with fewer evaluations.
Incorrect! Try again.
59A practitioner repeatedly chooses hyperparameters based on cross-validation results and reports the best cross-validation score as final performance. What is the primary methodological problem?
Model Improvement (Cross Validation, Hyperparameter Tuning, Grid Search, Random Search)
Hard
A.Hyperparameters are learned from target values by definition
B.Cross-validation cannot compare hyperparameters
C.The training set is always smaller than the test set
D.The validation results have become part of the model-selection process
Correct Answer: The validation results have become part of the model-selection process
Explanation:
Repeatedly optimizing against the same cross-validation results can overfit those results. A separate untouched test set, or nested cross-validation, is needed for an unbiased performance estimate.
Incorrect! Try again.
60A feature attribution method assigns positive contributions to several correlated features. Why should these individual attributions be interpreted cautiously?
Introduction to Explainable AI
Hard
A.Positive contributions prove that removing each feature improves accuracy
B.Correlation guarantees that every attribution is causally correct
C.Correlated features can share or redistribute credit across equivalent explanations
D.Attributions are always independent of the model's training data
Correct Answer: Correlated features can share or redistribute credit across equivalent explanations
Explanation:
When features contain overlapping information, attribution methods may distribute importance among them in different ways. An attribution describes model behavior, not necessarily an independent causal effect.
Incorrect! Try again.
Did this save you a night before the exam?
LPU Notes is free, and it stays free. Ads cover part of the server bill.
The rest comes out of a student's own pocket: the domain, the storage,
and keeping the site up through the weeks everyone needs it at once.
The payment button didn't load. An ad blocker or a filtered network is the usual reason.
to try again.
Nothing here is ever locked, and nothing unlocks. Chip in only if it was worth it.
What it pays for →