Machine Learning MCQ : Test 4

Explore this diverse selection of multiple-choice questions (MCQs) designed for various examinations. Machine Learning MCQ : Test 4 focuses on essential aspects of the subject, ensuring comprehensive preparation across different categories and fields of study to enhance your knowledge and readiness. The right answers for each question is provided next to respective questions for your convenience, you can either attend the test or dirtectly access the right answers by clicking the show correct answer button.

Each correct answer earns 1 mark, while each incorrect answer deducts 0.3 marks.

1. What is the purpose of the softmax function?

A. To reduce data dimensions

B. To transform values into probabilities

C. To cluster data points

D. To increase model complexity

2. Which method is used for dimensionality reduction?

A. K-means

B. PCA

C. Decision tree

D. Logistic regression

3. What is the role of an activation function in a neural network?

A. To normalize inputs

B. To introduce non-linearity

C. To reduce data size

D. To increase data dimensions

4. Which evaluation metric is used for regression?

A. Accuracy

B. Precision

C. Mean absolute error (MAE)

D. Recall

5. What is the purpose of k-fold cross-validation?

A. To increase data size

B. To reduce overfitting

C. To evaluate model performance

D. To cluster data points

6. How does ensemble learning improve model performance?

A. By using a single model

B. By combining multiple models

C. By reducing data size

D. By increasing data dimensions

7. Which technique is used for image classification?

A. Decision tree

B. Logistic regression

C. Convolutional neural network (CNN)

D. K-means

8. What is the role of a loss function in machine learning?

A. To increase data dimensions

B. To measure the difference between predicted and actual values

C. To cluster data points

D. To normalize inputs

9. Which algorithm is used for sentiment analysis?

A. K-means

B. Decision tree

C. Logistic regression

D. Apriori

10. What is the purpose of a learning curve?

A. To reduce data size

B. To visualize model performance over time

C. To increase model complexity

D. To cluster data points

11. How does a convolutional neural network (CNN) process images?

A. By reducing data dimensions

B. By using convolutional layers to detect features

C. By clustering data points

D. By increasing data size

12. Which method is used for dimensionality reduction?

A. Linear regression

B. K-means

C. Principal component analysis (PCA)

D. Decision tree

13. What is a confusion matrix?

A. A clustering method

B. A table used to evaluate the performance of a classification model

C. A type of neural network

D. A dimensionality reduction technique

14. What is the curse of dimensionality?

A. High computational cost with high-dimensional data

B. Lack of data in high dimensions

C. Reduced model complexity

D. Increased overfitting

15. Which of the following is a kernel trick in SVM?

A. Linear transformation

B. Mapping data to a higher dimension

C. Reducing the number of features

D. Increasing data size

16. What does the term 'bias-variance tradeoff' describe?

A. The balance between model simplicity and complexity

B. The tradeoff between training and testing error

C. The balance between data size and model performance

D. The tradeoff between supervised and unsupervised learning

17. How does L1 regularization work?

A. By adding the absolute value of the coefficients to the cost function

B. By adding the squared value of the coefficients to the cost function

C. By reducing the number of features

D. By increasing the data size

18. Which optimization algorithm is used for training deep neural networks?

A. Gradient descent

B. Genetic algorithm

C. Simulated annealing

D. Particle swarm optimization

19. Which of the following is a common problem with high-dimensional data?

A. Underfitting

B. Overfitting

C. High accuracy

D. Low bias

20. What is the purpose of dropout in neural networks?

A. To reduce model complexity

B. To prevent overfitting by randomly dropping units during training

C. To increase model interpretability

D. To simplify the network

21. How does a convolutional neural network (CNN) differ from a fully connected network?

A. CNN uses local connections and weight sharing

B. CNN uses global connections and no weight sharing

C. CNN has no hidden layers

D. CNN only uses linear activation functions

22. Which method is used to evaluate the performance of a clustering algorithm?

A. Precision

B. Recall

C. Silhouette score

D. F1 score

23. What is the role of the learning rate in gradient descent?

A. To control the step size during optimization

B. To reduce the number of features

C. To increase the data size

D. To transform the data

24. Which algorithm is used for topic modeling in NLP?

A. K-means clustering

B. Decision tree

C. Latent Dirichlet Allocation (LDA)

D. Logistic regression

25. How does the AdaBoost algorithm work?

A. By combining multiple weak classifiers to form a strong classifier

B. By reducing data dimensions

C. By clustering data points

D. By increasing model complexity

Question Navigation

Take Test