Classification Algorithm-2 MCQs – Class 12 Data Science (844)
Ace your CBSE boards with these Classification Algorithm-2 MCQs — fully exam-oriented, covering all topics, designed to boost concepts and maximize your score.
Q1. The K-Nearest Neighbors (K-NN) algorithm belongs to which category of machine learning?
a) Unsupervised Learning
b) Supervised Learning
c) Reinforcement Learning
d) Semi-supervised Learning
Q2. K-NN works on the principle that:
a) Similar things exist close to each other
b) Dissimilar things exist close to each other
c) All data points are equally distant
d) Data must always be normally distributed
Q3. In K-NN, what does ‘K’ represent?
a) The total number of classes in the dataset
b) The number of points checked before a majority vote
c) The number of features in the dataset
d) The number of training iterations
Q4. Why do we usually prefer odd values of K in K-NN?
a) To reduce memory usage
b) To help the algorithm take a clear majority vote and avoid ties
c) Odd values always give higher accuracy
d) It reduces the number of features required
Q5. A large value of K in K-NN:
a) Increases the effect of outliers
b) Reduces the effect of outliers but causes the decision surface to generalize
c) Has no effect on the decision surface
d) Removes the need for majority voting
Q6. K-NN is described as a “lazy learner” because:
a) It takes too long to code
b) It does not have a specific training phase and uses all training data during classification
c) It only works on small datasets
d) It ignores most of the training data
Q7. K-NN is called a non-parametric algorithm because:
a) It requires no data at all
b) It does not assume anything about the distribution of the data
c) It only works with categorical data
d) It cannot be tuned
Q8. Which of the following is the only hyperparameter of the K-NN algorithm?
a) Learning rate
b) K
c) Number of epochs
d) Depth of tree
Q9. Which of the following is an advantage of K-NN?
a) High interpretability and ease of understanding
b) Requires heavy training time
c) Poor performance on multi-class problems
d) Cannot handle regression problems
Q10. Which of the following is a major disadvantage of K-NN when the dataset becomes very large?
a) It becomes faster
b) It requires fewer calculations
c) Its speed declines significantly because it evaluates the data points during prediction
d) It automatically reduces the dataset
Q11. Unlike parametric models such as linear regression, K-NN:
a) Makes strong assumptions about data distribution
b) Makes no assumptions about data distribution
c) Cannot be used for prediction
d) Requires normally distributed data
Q12. What is a major drawback of K-NN as the dataset grows larger?
a) It becomes faster
b) It becomes slow and memory inefficient
c) It stops working for classification
d) It no longer needs distance calculation
Q13. Why does K-NN need to store the entire dataset in memory during prediction?
a) Because it has no training phase and evaluates all points at prediction time
b) Because it deletes data after each prediction
c) Because it only stores the mean of the data
d) Because it compresses the data first
Q14. K-NN gives poor accuracy on datasets that are:
a) Balanced
b) Imbalanced, biasing predictions toward the majority class
c) Very small
d) Free of outliers
Q15. K-NN is highly sensitive to:
a) The number of classes
b) Outliers, especially near the decision boundary
c) The programming language used
d) The name of the features
Q16. A K-NN model works well with two input variables but struggles when hundreds of variables are introduced. Which problem does this illustrate?
a) Overfitting due to validation data
b) Curse of dimensionality
c) Majority voting
d) Lazy learning
Q17. As a thumb rule, if unsure which value of k to choose for cross-validation, one can take:
a) k = 1
b) k = 3
c) k = 10
d) k = 100
Q18. According to the chapter, cross-validation can also be used with K-NN to:
a) Determine the value of K
b) Remove the need for data
c) Convert K-NN into regression
d) Eliminate all high-dimensional variables
Q19. Which sequence correctly represents the basic steps of cross-validation described in the chapter?
a) Test → Reserve → Train
b) Train → Reserve → Test
c) Reserve validation data → Train on remaining data → Test on validation data
d) Reserve all data → Test → Train
Q20. Assertion (A): K-NN is a lazy learner.
Reason (R): K-NN does not have a specific training phase and uses the training data while performing classification.
a) Both A and R are true, and R is the correct explanation of A
b) Both A and R are true, but R is NOT the correct explanation of A
c) A is true but R is false
d) A is false but R is true
Q21. Assertion (A): K-NN can be affected by imbalanced datasets.
Reason (R): When one class forms the majority of the training data, K-NN may become biased towards that class.
a) Both A and R are true, and R is the correct explanation of A
b) Both A and R are true, but R is NOT the correct explanation of A
c) A is true but R is false
d) A is false but R is true
Q22. Assertion (A): Cross-validation reserves a portion of the dataset and does not use it for training the model.
Reason (R): The reserved portion is used to test the resulting model before finalizing it.
a) Both A and R are true, and R is the correct explanation of A
b) Both A and R are true, but R is NOT the correct explanation of A
c) A is true but R is false
d) A is false but R is true
