Disclaimer:
Educational Use Only:
The website links provided are to be used strictly for educational and knowledge-gaining purposes only.
No Misuse:
Do not misuse the links or the information contained therein for any commercial, illegal, or unethical activities.
Respect for Websites and Businesses:
Respect the intellectual property and business interests of the websites and their owners. Do not attempt to alter, copy, or redistribute content without proper authorization.
Acknowledge Value of Knowledge:
Recognize and value the knowledge and resources shared by these websites. Use the information responsibly and ethically.
No Endorsement:
The provision of these links does not imply endorsement of the linked sites, their content, or their products and services. Users are responsible for evaluating the suitability and accuracy of the information.
User Responsibility:
By accessing these links, you agree to use the information at your own risk and to comply with all applicable laws and website terms of use.
Overall MCQs for Reference
- Which metric is commonly used to evaluate classification models in the presence of class imbalance?
a. R-squared
b. F1-score
c. Accuracy
d. Mean Squared Error (MSE)
Answer: b. F1-scoreinterviewkickstart+2 - What does ‘Precision’ measure in classification tasks?
a. The proportion of actual positives correctly identified
b. The proportion of predicted positives that are correct
c. The proportion of actual negatives correctly identified
d. The overall classification accuracy
Answer: b. The proportion of predicted positives that are correctlearncbse - Which evaluation metric combines precision and recall into a single value?
a. ROC-AUC
b. F1-score
c. MAE
d. R-squared
Answer: b. F1-scoreai-cbse+1 - Which metric is best for measuring regression model performance?
a. Accuracy
b. Mean Squared Error (MSE)
c. F1-score
d. Recall
Answer: b. Mean Squared Error (MSE)igmguru - What does ‘Recall’ specifically measure in a binary classifier?
a. Fraction of true positives among predicted positives
b. Fraction of true positives among actual positives
c. Fraction of true negatives among actual negatives
d. Proportion of all correct predictions
Answer: b. Fraction of true positives among actual positivesai-cbse+1 - A high ROC-AUC score indicates what about a model’s performance?
a. It is underfitting
b. It can distinguish classes well
c. Predictions are random
d. It is overfitting
Answer: b. It can distinguish classes wellai-cbse - Which result represents a perfect F1-score?
a. 0.0
b. 0.5
c. 1.0
d. 2.0
Answer: c. 1.0learncbse - What does the confusion matrix NOT contain?
a. True Positives
b. False Negatives
c. Precision
d. False Positives
Answer: c. Precisionai-cbse - Which metric is least affected by outliers in regression?
a. Mean Squared Error
b. Root Mean Squared Error
c. Mean Absolute Error
d. R-squared
Answer: c. Mean Absolute Errorsanfoundry - What is indicated if a model has high accuracy but low recall?
a. It predicts too many true negatives
b. It misses many true positives
c. It predicts too many false positives
d. It is highly sensitive
Answer: b. It misses many true positiveslearncbse+1 - For multiclass classification, which metric is most informative?
a. Confusion Matrix
b. Mean Squared Error
c. R-squared
d. ROC-AUC
Answer: a. Confusion Matrixsanfoundry+1 - What is the key purpose of feature importance ranking in model analysis?
a. Increase bias
b. Reduce variance
c. Explain predictions
d. Avoid overfitting
Answer: c. Explain predictionsinterviewkickstart+1 - If a regression model’s R-squared is 0.97, what does it mean?
a. The model explains 97% of the variance
b. The error is 97%
c. The accuracy is 97%
d. The recall is 97%
Answer: a. The model explains 97% of the varianceigmguru+1 - The F1-score is most useful when:
a. Data is balanced
b. Data is imbalanced
c. Only accuracy matters
d. The model has high bias
Answer: b. Data is imbalancedinterviewkickstart+2 - Which measure calculates the average error magnitude?
a. Accuracy
b. MAE (Mean Absolute Error)
c. ROC-AUC
d. Recall
Answer: b. MAE (Mean Absolute Error)sanfoundry - What does overfitting typically result in?
a. High training error, low test error
b. High test error, low training error
c. Low error in both
d. Equal error rates
Answer: b. High test error, low training errorigmguru+1 - Which technique helps to evaluate model generalization?
a. Regularization
b. Cross-validation
c. Random forest
d. Feature selection
Answer: b. Cross-validationinterviewkickstart+2 - In the confusion matrix, which term refers to incorrectly predicted positive cases?
a. True Positive
b. False Negative
c. False Positive
d. True Negative
Answer: c. False Positiveai-cbse - What is the main aim of a model validation set?
a. Training the model
b. Testing performance on seen data
c. Assessing performance on unseen data
d. Increasing overfitting
Answer: c. Assessing performance on unseen dataexamveda+1 - Which metric would be best for segmentation tasks in image analysis?
a. Dice coefficient
b. Accuracy
c. MAE
d. Precision
Answer: a. Dice coefficientigmguru - Which term describes the difference between predicted and actual values in regression?
a. Error
b. Recall
c. Precision
d. Specificity
Answer: a. Errorsanfoundry+1 - Root Mean Squared Error (RMSE) is preferable over MAE when:
a. There are many outliers
b. Penalizing large errors is important
c. Precision is required
d. The model is for classification
Answer: b. Penalizing large errors is importantsanfoundry - Which is NOT part of AI model performance analysis?
a. Evaluation metrics
b. Data preprocessing
c. Feature engineering
d. Training loss computation
Answer: c. Feature engineeringinterviewkickstart+1 - What does specificity measure?
a. True positives among all positives
b. True negatives among all negatives
c. False positives among all negatives
d. False negatives among all positives
Answer: b. True negatives among all negativeslearncbse+1 - Model robustness is best checked by:
a. Using only clean data
b. Testing against noisy data
c. Increasing layers
d. Removing features
Answer: b. Testing against noisy dataigmguru+1 - Which metric may mislead when classes are skewed?
a. Accuracy
b. F1-score
c. Precision
d. Recall
Answer: a. Accuracyinterviewkickstart+1 - What is the best method for tuning hyperparameters?
a. Cross-validation
b. Increasing feature set
c. Reducing training epochs
d. Lowering learning rate
Answer: a. Cross-validationexamveda+1 - Which of the following is used to avoid overfitting?
a. Increasing the number of features
b. Data augmentation
c. Increasing model complexity
d. Removing validation set
Answer: b. Data augmentationigmguru - “Bias-Variance Tradeoff” affects which aspect of AI performance?
a. Model complexity
b. Accuracy only
c. Model interpretability
d. Prediction speed
Answer: a. Model complexityinterviewkickstart+1 - Which action improves recall?
a. Lowering threshold for classification
b. Increasing threshold for classification
c. Removing validation data
d. Reducing features
Answer: a. Lowering threshold for classificationlearncbse+1 - Which metric describes the fraction of negative samples correctly identified?
a. Precision
b. Recall
c. Specificity
d. F1-score
Answer: c. Specificityai-cbse+1 - Which performance metric is used for clustering analysis?
a. Silhouette score
b. MAE
c. Precision
d. ROC-AUC
Answer: a. Silhouette scoresanfoundry+1 - Which method best evaluates a model’s ability to handle real-world data changes?
a. Static validation
b. Dynamic testing
c. Regularization
d. Feature engineering
Answer: b. Dynamic testingai-cbse+1 - The confusion matrix is most useful for evaluating:
a. Regression tasks
b. Classification tasks
c. Clustering analysis
d. Time series prediction
Answer: b. Classification taskssanfoundry+1 - Which metric combines TPR and FPR across thresholds?
a. Accuracy
b. ROC-AUC
c. MAE
d. F1-score
Answer: b. ROC-AUCai-cbse - Which is a drawback of using only accuracy to measure performance?
a. It ignores data imbalance
b. It is too mathematical
c. It requires more computation
d. It is widely accepted
Answer: a. It ignores data imbalanceinterviewkickstart+1 - Which evaluation method checks for overfitting?
a. Comparing train and test errors
b. Increasing epoch count
c. Adding features
d. Lowering learning rate
Answer: a. Comparing train and test errorsigmguru+1 - What does “sensitivity” in performance metrics refer to?
a. Precision
b. Recall
c. Specificity
d. Accuracy
Answer: b. Recalllearncbse+1 - Which approach best measures performance on unseen scenarios?
a. K-fold cross-validation
b. Single split validation
c. Using training set only
d. Ignoring validation
Answer: a. K-fold cross-validationexamveda+1 - A high FPR (False Positive Rate) most adversely affects:
a. Spam detection systems
b. Regression models
c. Clustering analysis
d. Feature selection
Answer: a. Spam detection systemsai-cbse - Adjusting which parameter will typically change the balance between precision and recall?
a. Threshold
b. Learning rate
c. Epochs
d. Feature count
Answer: a. Thresholdlearncbse+1 - Which method estimates confidence in predictions?
a. Bootstrap resampling
b. Regularization
c. Silhouette score
d. MAE
Answer: a. Bootstrap resamplingigmguru - For time-series forecasting, which metric is commonly used?
a. MAE
b. F1-score
c. ROC-AUC
d. Silhouette score
Answer: a. MAEsanfoundry+1 - Negative predictive value measures:
a. True negatives among predicted negatives
b. True positives among predicted negatives
c. False positives among predicted negatives
d. False negatives among predicted negatives
Answer: a. True negatives among predicted negativeslearncbse+1 - Which metric is least relevant for regression tasks?
a. R-squared
b. RMSE
c. Precision
d. MAE
Answer: c. Precisionsanfoundry+1 - What is the impact of increasing model complexity in limited data?
a. Increased overfitting risk
b. Improved generalization
c. Decreased bias
d. Lower test error
Answer: a. Increased overfitting riskinterviewkickstart+1 - Which metric would be most suitable when false negatives are critical?
a. Precision
b. Recall
c. Specificity
d. ROC-AUC
Answer: b. Recalllearncbse+1 - Which is a direct method for comparing two models’ performance?
a. Paired t-test on errors
b. Removing features
c. Lowering epoch number
d. Ignoring validation
Answer: a. Paired t-test on errorstandfonline+1 - If a model’s test error increases, but training error decreases, this likely means:
a. Overfitting is occurring
b. Underfitting is occurring
c. Training data lacks variance
d. Validation set is small
Answer: a. Overfitting is occurringigmguru+1 - What does ROC curve plot?
a. Precision vs. Recall
b. Sensitivity vs. Specificity
c. True Positive Rate vs. False Positive Rate
d. Error Rate vs. Model Complexity
Answer: c. True Positive Rate vs. False Positive Rateai-cbse
- https://interviewkickstart.com/blogs/articles/machine-learning-mcqs-answers
- https://www.ai-cbse.com/question-bank/unit-7-evaluation/
- https://www.igmguru.com/blog/machine-learning-mcqs
- https://www.learncbse.in/evaluation-class-10-mcq/
- https://www.sanfoundry.com/1000-machine-learning-questions-answers/
- https://www.examveda.com/data-science/practice-mcq-question-on-model-evaluation-and-validation/?page=3
- https://www.tandfonline.com/doi/full/10.1080/02602938.2023.2299059
- https://www.sanfoundry.com/artificial-intelligence-questions-answers/
- https://testbook.com/objective-questions/mcq-on-artificial-intelligence–5eea6a0d39140f30f369e30e
- https://www.interviewbit.com/artificial-intelligence-mcq/
- https://www.igmguru.com/blog/artificial-intelligence-mcqs
- https://pmc.ncbi.nlm.nih.gov/articles/PMC12158464/
- https://testbook.com/objective-questions/mcq-on-artificial-intelligence–5eea6a0e39140f30f369e51f
- https://www.scribd.com/document/737189591/Artificial-Intelligence-MCQ-unit1-2
- https://www.sciencedirect.com/science/article/pii/S2665944125000227
- https://www.sanfoundry.com/artificial-intelligence-mcqs-problem-solving-1/
- https://www.scribd.com/document/713325576/100-MCQ-QUESTIONS-FOR-PRACTICE