AI/ML Fundamentals MCQ Questions with Answers – Page 2 (Latest 2026)

Q51. Which statement about cross-validation is most accurate?

A. Single train-test split only. B. Estimating generalization via multiple train/val splits. C. Hand-tuning per epoch. D. Training on the test set.

Select an answer to check.

Answer: Estimating generalization via multiple train/val splits.

Here, Estimating generalization via multiple train/val splits. is the right choice. K-fold is most common. It aligns directly with what the question asks about which statement about cross-validation is most accurate. Competing choices sound plausible, but they miss the key condition.

Q52. How is cross-validation best characterized?

A. Estimating generalization via multiple train/val splits. B. Hand-tuning per epoch. C. Training on the test set. D. Single train-test split only.

Select an answer to check.

Answer: Estimating generalization via multiple train/val splits.

In this case, Estimating generalization via multiple train/val splits. is correct. K-fold is most common. It aligns directly with what the question asks about how is cross-validation best characterized. Competing choices sound plausible, but they miss the key condition.

Q53. Which option best describes gradient descent?

A. Iteratively updating params opposite to the gradient of loss. B. Random parameter perturbation only. C. A vector store query. D. An exact closed-form solution.

Select an answer to check.

Answer: Iteratively updating params opposite to the gradient of loss.

The best option here is Iteratively updating params opposite to the gradient of loss.. Backbone of most ML training. It aligns directly with what the question asks about which option best describes gradient descent. Competing choices sound plausible, but they miss the key condition.

Q54. What is the primary purpose of gradient descent?

A. A vector store query. B. Random parameter perturbation only. C. An exact closed-form solution. D. Iteratively updating params opposite to the gradient of loss.

Select an answer to check.

Answer: Iteratively updating params opposite to the gradient of loss.

For this question, Iteratively updating params opposite to the gradient of loss. is correct. Backbone of most ML training. It aligns directly with what the question asks about what is the primary purpose of gradient descent. Competing choices sound plausible, but they miss the key condition.

Q55. Which statement about gradient descent is most accurate?

A. Random parameter perturbation only. B. A vector store query. C. Iteratively updating params opposite to the gradient of loss. D. An exact closed-form solution.

Select an answer to check.

Answer: Iteratively updating params opposite to the gradient of loss.

Iteratively updating params opposite to the gradient of loss. is the correct answer here. Backbone of most ML training. It aligns directly with what the question asks about which statement about gradient descent is most accurate. Competing choices sound plausible, but they miss the key condition.

Q56. How is gradient descent best characterized?

A. An exact closed-form solution. B. Iteratively updating params opposite to the gradient of loss. C. Random parameter perturbation only. D. A vector store query.

Select an answer to check.

Answer: Iteratively updating params opposite to the gradient of loss.

Here, Iteratively updating params opposite to the gradient of loss. is the right choice. Backbone of most ML training. This matches the core idea being tested around how is gradient descent best characterized. Competing choices sound plausible, but they miss the key condition.

Q57. Which option best describes learning rate?

A. Number of training samples. B. Number of features. C. Step size for gradient updates. D. Number of layers.

Select an answer to check.

Answer: Step size for gradient updates.

In this case, Step size for gradient updates. is correct. Critical hyperparameter. This matches the core idea being tested around which option best describes learning rate. Competing choices sound plausible, but they miss the key condition.

Q58. What is the primary purpose of learning rate?

A. Number of training samples. B. Number of features. C. Step size for gradient updates. D. Number of layers.

Select an answer to check.

Answer: Step size for gradient updates.

The best option here is Step size for gradient updates.. Critical hyperparameter. This matches the core idea being tested around what is the primary purpose of learning rate. Competing choices sound plausible, but they miss the key condition.

Q59. Which statement about learning rate is most accurate?

A. Number of features. B. Step size for gradient updates. C. Number of training samples. D. Number of layers.

Select an answer to check.

Answer: Step size for gradient updates.

For this question, Step size for gradient updates. is correct. Critical hyperparameter. This matches the core idea being tested around which statement about learning rate is most accurate. Competing choices sound plausible, but they miss the key condition.

Q60. How is learning rate best characterized?

A. Number of layers. B. Number of features. C. Number of training samples. D. Step size for gradient updates.

Select an answer to check.

Answer: Step size for gradient updates.

Step size for gradient updates. is the correct answer here. Critical hyperparameter. This matches the core idea being tested around how is learning rate best characterized. Competing choices sound plausible, but they miss the key condition.

Q61. Which option best describes loss function?

A. A model layer. B. An accelerator type. C. Quantifies prediction error to be minimized. D. An evaluation metric only.

Select an answer to check.

Answer: Quantifies prediction error to be minimized.

Here, Quantifies prediction error to be minimized. is the right choice. Differentiable for gradient-based opt. That is exactly the concept behind which option best describes loss function in this context. Competing choices sound plausible, but they miss the key condition.

Q62. What is the primary purpose of loss function?

A. An evaluation metric only. B. An accelerator type. C. Quantifies prediction error to be minimized. D. A model layer.

Select an answer to check.

Answer: Quantifies prediction error to be minimized.

In this case, Quantifies prediction error to be minimized. is correct. Differentiable for gradient-based opt. That is exactly the concept behind what is the primary purpose of loss function in this context. Competing choices sound plausible, but they miss the key condition.

Q63. Which statement about loss function is most accurate?

A. An evaluation metric only. B. A model layer. C. An accelerator type. D. Quantifies prediction error to be minimized.

Select an answer to check.

Answer: Quantifies prediction error to be minimized.

The best option here is Quantifies prediction error to be minimized.. Differentiable for gradient-based opt. That is exactly the concept behind which statement about loss function is most accurate in this context. Competing choices sound plausible, but they miss the key condition.

Q64. How is loss function best characterized?

A. Quantifies prediction error to be minimized. B. An accelerator type. C. An evaluation metric only. D. A model layer.

Select an answer to check.

Answer: Quantifies prediction error to be minimized.

For this question, Quantifies prediction error to be minimized. is correct. Differentiable for gradient-based opt. That is exactly the concept behind how is loss function best characterized in this context. Competing choices sound plausible, but they miss the key condition.

Q65. Which option best describes classification?

A. Predicting a continuous value. B. Predicting a discrete category. C. Generating images. D. Clustering unlabeled data.

Select an answer to check.

Answer: Predicting a discrete category.

Predicting a discrete category. is the correct answer here. Examples: spam detection. That is exactly the concept behind which option best describes classification in this context. Competing choices sound plausible, but they miss the key condition.

Q66. What is the primary purpose of classification?

A. Clustering unlabeled data. B. Predicting a discrete category. C. Predicting a continuous value. D. Generating images.

Select an answer to check.

Answer: Predicting a discrete category.

Here, Predicting a discrete category. is the right choice. Examples: spam detection. It fits the requirement in the prompt about what is the primary purpose of classification. Competing choices sound plausible, but they miss the key condition.

Q67. Which statement about classification is most accurate?

A. Predicting a continuous value. B. Generating images. C. Clustering unlabeled data. D. Predicting a discrete category.

Select an answer to check.

Answer: Predicting a discrete category.

In this case, Predicting a discrete category. is correct. Examples: spam detection. It fits the requirement in the prompt about which statement about classification is most accurate. Competing choices sound plausible, but they miss the key condition.

Q68. How is classification best characterized?

A. Clustering unlabeled data. B. Predicting a continuous value. C. Generating images. D. Predicting a discrete category.

Select an answer to check.

Answer: Predicting a discrete category.

The best option here is Predicting a discrete category.. Examples: spam detection. It fits the requirement in the prompt about how is classification best characterized. Competing choices sound plausible, but they miss the key condition.

Q69. Which option best describes regression?

A. Generating text. B. Predicting a category label. C. Predicting a continuous value. D. Discovering clusters.

Select an answer to check.

Answer: Predicting a continuous value.

For this question, Predicting a continuous value. is correct. Examples: house price prediction. It fits the requirement in the prompt about which option best describes regression. Competing choices sound plausible, but they miss the key condition.

Q70. What is the primary purpose of regression?

A. Predicting a continuous value. B. Generating text. C. Predicting a category label. D. Discovering clusters.

Select an answer to check.

Answer: Predicting a continuous value.

Predicting a continuous value. is the correct answer here. Examples: house price prediction. It fits the requirement in the prompt about what is the primary purpose of regression. Competing choices sound plausible, but they miss the key condition.

Q71. Which statement about regression is most accurate?

A. Discovering clusters. B. Predicting a category label. C. Predicting a continuous value. D. Generating text.

Select an answer to check.

Answer: Predicting a continuous value.

Here, Predicting a continuous value. is the right choice. Examples: house price prediction. This is the most accurate statement for which statement about regression is most accurate. Competing choices sound plausible, but they miss the key condition.

Q72. How is regression best characterized?

A. Predicting a category label. B. Predicting a continuous value. C. Generating text. D. Discovering clusters.

Select an answer to check.

Answer: Predicting a continuous value.

In this case, Predicting a continuous value. is correct. Examples: house price prediction. This is the most accurate statement for how is regression best characterized. Competing choices sound plausible, but they miss the key condition.

Q73. Which option best describes accuracy?

A. Mean squared error of regression. B. A retrieval recall metric. C. A loss function for training. D. Fraction of correct predictions over all predictions.

Select an answer to check.

Answer: Fraction of correct predictions over all predictions.

The best option here is Fraction of correct predictions over all predictions.. Misleading on imbalanced data. This is the most accurate statement for which option best describes accuracy. Competing choices sound plausible, but they miss the key condition.

Q74. What is the primary purpose of accuracy?

A. A loss function for training. B. Mean squared error of regression. C. A retrieval recall metric. D. Fraction of correct predictions over all predictions.

Select an answer to check.

Answer: Fraction of correct predictions over all predictions.

For this question, Fraction of correct predictions over all predictions. is correct. Misleading on imbalanced data. This is the most accurate statement for what is the primary purpose of accuracy. Competing choices sound plausible, but they miss the key condition.

Q75. Which statement about accuracy is most accurate?

A. Mean squared error of regression. B. A loss function for training. C. Fraction of correct predictions over all predictions. D. A retrieval recall metric.

Select an answer to check.

Answer: Fraction of correct predictions over all predictions.

Fraction of correct predictions over all predictions. is the correct answer here. Misleading on imbalanced data. This is the most accurate statement for which statement about accuracy is most accurate. Competing choices sound plausible, but they miss the key condition.

Q76. How is accuracy best characterized?

A. Fraction of correct predictions over all predictions. B. A loss function for training. C. A retrieval recall metric. D. Mean squared error of regression.

Select an answer to check.

Answer: Fraction of correct predictions over all predictions.

Here, Fraction of correct predictions over all predictions. is the right choice. Misleading on imbalanced data. It aligns directly with what the question asks about how is accuracy best characterized. The remaining choices fail because they don’t satisfy the full definition.

Q77. Which option best describes precision?

A. Overall accuracy. B. TP / (TP + FN); recall. C. TP / (TP + FP); fraction of predicted positives that are correct. D. Mean error.

Select an answer to check.

Answer: TP / (TP + FP); fraction of predicted positives that are correct.

In this case, TP / (TP + FP); fraction of predicted positives that are correct. is correct. Penalizes false positives. It aligns directly with what the question asks about which option best describes precision. The remaining choices fail because they don’t satisfy the full definition.

Q78. What is the primary purpose of precision?

A. TP / (TP + FN); recall. B. Overall accuracy. C. TP / (TP + FP); fraction of predicted positives that are correct. D. Mean error.

Select an answer to check.

Answer: TP / (TP + FP); fraction of predicted positives that are correct.

The best option here is TP / (TP + FP); fraction of predicted positives that are correct.. Penalizes false positives. It aligns directly with what the question asks about what is the primary purpose of precision. The remaining choices fail because they don’t satisfy the full definition.

Q79. Which statement about precision is most accurate?

A. Overall accuracy. B. Mean error. C. TP / (TP + FP); fraction of predicted positives that are correct. D. TP / (TP + FN); recall.

Select an answer to check.

Answer: TP / (TP + FP); fraction of predicted positives that are correct.

For this question, TP / (TP + FP); fraction of predicted positives that are correct. is correct. Penalizes false positives. It aligns directly with what the question asks about which statement about precision is most accurate. The remaining choices fail because they don’t satisfy the full definition.

Q80. How is precision best characterized?

A. Mean error. B. Overall accuracy. C. TP / (TP + FN); recall. D. TP / (TP + FP); fraction of predicted positives that are correct.

Select an answer to check.

Answer: TP / (TP + FP); fraction of predicted positives that are correct.

TP / (TP + FP); fraction of predicted positives that are correct. is the correct answer here. Penalizes false positives. It aligns directly with what the question asks about how is precision best characterized. The remaining choices fail because they don’t satisfy the full definition.

Q81. Which option best describes recall?

A. TP / (TP + FP); precision. B. Overall accuracy. C. TP / (TP + FN); fraction of actual positives correctly identified. D. Mean error.

Select an answer to check.

Answer: TP / (TP + FN); fraction of actual positives correctly identified.

Here, TP / (TP + FN); fraction of actual positives correctly identified. is the right choice. Penalizes false negatives. This matches the core idea being tested around which option best describes recall. The remaining choices fail because they don’t satisfy the full definition.

Q82. What is the primary purpose of recall?

A. TP / (TP + FP); precision. B. Overall accuracy. C. Mean error. D. TP / (TP + FN); fraction of actual positives correctly identified.

Select an answer to check.

Answer: TP / (TP + FN); fraction of actual positives correctly identified.

In this case, TP / (TP + FN); fraction of actual positives correctly identified. is correct. Penalizes false negatives. This matches the core idea being tested around what is the primary purpose of recall. The remaining choices fail because they don’t satisfy the full definition.

Q83. Which statement about recall is most accurate?

A. Overall accuracy. B. TP / (TP + FN); fraction of actual positives correctly identified. C. TP / (TP + FP); precision. D. Mean error.

Select an answer to check.

Answer: TP / (TP + FN); fraction of actual positives correctly identified.

The best option here is TP / (TP + FN); fraction of actual positives correctly identified.. Penalizes false negatives. This matches the core idea being tested around which statement about recall is most accurate. The remaining choices fail because they don’t satisfy the full definition.

Q84. How is recall best characterized?

A. TP / (TP + FP); precision. B. Overall accuracy. C. TP / (TP + FN); fraction of actual positives correctly identified. D. Mean error.

Select an answer to check.

Answer: TP / (TP + FN); fraction of actual positives correctly identified.

For this question, TP / (TP + FN); fraction of actual positives correctly identified. is correct. Penalizes false negatives. This matches the core idea being tested around how is recall best characterized. The remaining choices fail because they don’t satisfy the full definition.

Q85. Which option best describes F1 score?

A. Harmonic mean of precision and recall. B. Arithmetic mean of accuracy and recall. C. Median of confusion matrix entries. D. Geometric mean of MSE and MAE.

Select an answer to check.

Answer: Harmonic mean of precision and recall.

Harmonic mean of precision and recall. is the correct answer here. Useful with class imbalance. This matches the core idea being tested around which option best describes f1 score. The remaining choices fail because they don’t satisfy the full definition.

Q86. What is the primary purpose of F1 score?

A. Harmonic mean of precision and recall. B. Arithmetic mean of accuracy and recall. C. Geometric mean of MSE and MAE. D. Median of confusion matrix entries.

Select an answer to check.

Answer: Harmonic mean of precision and recall.

Here, Harmonic mean of precision and recall. is the right choice. Useful with class imbalance. That is exactly the concept behind what is the primary purpose of f1 score in this context. The remaining choices fail because they don’t satisfy the full definition.

Q87. Which statement about F1 score is most accurate?

A. Harmonic mean of precision and recall. B. Arithmetic mean of accuracy and recall. C. Median of confusion matrix entries. D. Geometric mean of MSE and MAE.

Select an answer to check.

Answer: Harmonic mean of precision and recall.

In this case, Harmonic mean of precision and recall. is correct. Useful with class imbalance. That is exactly the concept behind which statement about f1 score is most accurate in this context. The remaining choices fail because they don’t satisfy the full definition.

Q88. How is F1 score best characterized?

A. Harmonic mean of precision and recall. B. Arithmetic mean of accuracy and recall. C. Median of confusion matrix entries. D. Geometric mean of MSE and MAE.

Select an answer to check.

Answer: Harmonic mean of precision and recall.

The best option here is Harmonic mean of precision and recall.. Useful with class imbalance. That is exactly the concept behind how is f1 score best characterized in this context. The remaining choices fail because they don’t satisfy the full definition.

Q89. Which option best describes a confusion matrix?

A. A correlation matrix of features. B. Table of TP/FP/TN/FN by class. C. A model weight matrix. D. A covariance matrix of labels.

Select an answer to check.

Answer: Table of TP/FP/TN/FN by class.

For this question, Table of TP/FP/TN/FN by class. is correct. Drives many classification metrics. That is exactly the concept behind which option best describes a confusion matrix in this context. The remaining choices fail because they don’t satisfy the full definition.

Q90. What is the primary purpose of a confusion matrix?

A. A correlation matrix of features. B. Table of TP/FP/TN/FN by class. C. A covariance matrix of labels. D. A model weight matrix.

Select an answer to check.

Answer: Table of TP/FP/TN/FN by class.

Table of TP/FP/TN/FN by class. is the correct answer here. Drives many classification metrics. That is exactly the concept behind what is the primary purpose of a confusion in this context. The remaining choices fail because they don’t satisfy the full definition.

Q91. Which statement about a confusion matrix is most accurate?

A. A model weight matrix. B. A covariance matrix of labels. C. A correlation matrix of features. D. Table of TP/FP/TN/FN by class.

Select an answer to check.

Answer: Table of TP/FP/TN/FN by class.

Here, Table of TP/FP/TN/FN by class. is the right choice. Drives many classification metrics. It fits the requirement in the prompt about which statement about a confusion matrix is most. The remaining choices fail because they don’t satisfy the full definition.

Q92. How is a confusion matrix best characterized?

A. A covariance matrix of labels. B. Table of TP/FP/TN/FN by class. C. A model weight matrix. D. A correlation matrix of features.

Select an answer to check.

Answer: Table of TP/FP/TN/FN by class.

In this case, Table of TP/FP/TN/FN by class. is correct. Drives many classification metrics. It fits the requirement in the prompt about how is a confusion matrix best characterized. The remaining choices fail because they don’t satisfy the full definition.

Q93. Which option best describes ROC curve?

A. Precision vs recall curve. B. TPR vs FPR across thresholds. C. Loss vs epoch curve. D. Cumulative gain curve.

Select an answer to check.

Answer: TPR vs FPR across thresholds.

The best option here is TPR vs FPR across thresholds.. AUC summarizes ROC. It fits the requirement in the prompt about which option best describes roc curve. The remaining choices fail because they don’t satisfy the full definition.

Q94. What is the primary purpose of ROC curve?

A. Precision vs recall curve. B. TPR vs FPR across thresholds. C. Loss vs epoch curve. D. Cumulative gain curve.

Select an answer to check.

Answer: TPR vs FPR across thresholds.

For this question, TPR vs FPR across thresholds. is correct. AUC summarizes ROC. It fits the requirement in the prompt about what is the primary purpose of roc curve. The remaining choices fail because they don’t satisfy the full definition.

Q95. Which statement about ROC curve is most accurate?

A. Cumulative gain curve. B. TPR vs FPR across thresholds. C. Loss vs epoch curve. D. Precision vs recall curve.

Select an answer to check.

Answer: TPR vs FPR across thresholds.

TPR vs FPR across thresholds. is the correct answer here. AUC summarizes ROC. It fits the requirement in the prompt about which statement about roc curve is most accurate. The remaining choices fail because they don’t satisfy the full definition.

Q96. How is ROC curve best characterized?

A. TPR vs FPR across thresholds. B. Precision vs recall curve. C. Cumulative gain curve. D. Loss vs epoch curve.

Select an answer to check.

Answer: TPR vs FPR across thresholds.

Here, TPR vs FPR across thresholds. is the right choice. AUC summarizes ROC. This is the most accurate statement for how is roc curve best characterized. The remaining choices fail because they don’t satisfy the full definition.

Q97. Which option best describes PR curve?

A. TPR vs FPR curve (ROC). B. Variance vs bias. C. Precision vs recall across thresholds. D. Loss vs epoch.

Select an answer to check.

Answer: Precision vs recall across thresholds.

In this case, Precision vs recall across thresholds. is correct. Better than ROC under heavy imbalance. This is the most accurate statement for which option best describes pr curve. The remaining choices fail because they don’t satisfy the full definition.

Q98. What is the primary purpose of PR curve?

A. Loss vs epoch. B. Variance vs bias. C. TPR vs FPR curve (ROC). D. Precision vs recall across thresholds.

Select an answer to check.

Answer: Precision vs recall across thresholds.

The best option here is Precision vs recall across thresholds.. Better than ROC under heavy imbalance. This is the most accurate statement for what is the primary purpose of pr curve. The remaining choices fail because they don’t satisfy the full definition.

Q99. Which statement about PR curve is most accurate?

A. Precision vs recall across thresholds. B. Loss vs epoch. C. TPR vs FPR curve (ROC). D. Variance vs bias.

Select an answer to check.

Answer: Precision vs recall across thresholds.

For this question, Precision vs recall across thresholds. is correct. Better than ROC under heavy imbalance. This is the most accurate statement for which statement about pr curve is most accurate. The remaining choices fail because they don’t satisfy the full definition.

Q100. How is PR curve best characterized?

A. TPR vs FPR curve (ROC). B. Loss vs epoch. C. Variance vs bias. D. Precision vs recall across thresholds.

Select an answer to check.

Answer: Precision vs recall across thresholds.

Precision vs recall across thresholds. is the correct answer here. Better than ROC under heavy imbalance. This is the most accurate statement for how is pr curve best characterized. The remaining choices fail because they don’t satisfy the full definition.