Overview of Multinomial Naive Bayes Your Name: 1. What is one major disadvantage of certain algorithms regarding feature assumptions? A. They are always accurate regardless of the dataset. B. They require a large amount of training data. C. They assume complete independence of features. D. They are too complex to implement. 2. Why might irrelevant features reduce an algorithm's performance? A. They always enhance model accuracy. B. They only affect the training phase but not the testing phase. C. They introduce noise that complicates the learning process. D. They are essential for making predictions. 3. Which application is suitable for the discussed algorithms? A. Forecasting climate change effects. B. Real-time traffic prediction based on speed. C. Spam detection in email systems. D. Pricing strategy adjustments in real-time markets. 4. What is the purpose of smoothing techniques in the context of these algorithms? A. To speed up the training process. B. To create more features for better performance. C. To ensure all features are used regardless of relevance. D. To avoid zero probabilities for unseen words. 5. What is a recommended practice to enhance the performance of these algorithms? A. Focusing only on numerical data without any feature engineering. B. Using larger datasets and ignoring feature selection. C. Implementing random sampling for the training set. D. Integrating feature selection and preprocessing methods. 6. What is the main assumption of the Multinomial Naive Bayes algorithm regarding features? A. Features are independent of each other, given the class. B. Features are irrelevant to class determination. C. Features are continuously distributed. D. Features are dependent on each other. 7. What is the purpose of Laplace smoothing in the MNB algorithm? A. To simplify the computation of posterior probabilities. B. To increase the count of all features uniformly. C. To prevent zero probabilities during calculation. D. To decrease the overall vocabulary size. 8. In the context of MNB, which of the following correctly describes the calculation of class probabilities? A. P(Class|Document) is a ratio involving P(Document|Class) and P(Document). B. P(Class|Document) is always equal to 1. C. P(Class|Document) does not depend on the features of the document. D. P(Class|Document) is calculated as the product of prior probabilities. 9. Which type of data is Multinomial Naive Bayes particularly well-suited for? A. Continuous numerical data. B. Unstructured data without clear features. C. Categorical and continuous mixed data. D. Discrete data such as text documents. 10. How does MNB predict the class of a document? A. By random selection among classes. B. By maximizing the posterior probability given the document's features. C. By selecting the class with the lowest prior probability. D. By averaging the probabilities of all classes. 11. Which of the following is NOT an advantage of Multinomial Naive Bayes? A. Good performance on text classification tasks. B. Ease of implementation. C. Ability to handle non-discrete data effectively. D. Speed in calculations for large datasets. 12. Which formula represents Bayes' theorem as applied in MNB? A. P(Class|Document) = P(Document) + P(Class|Document) / P(Class) B. P(Class|Document) = P(Document|Class) + P(Class) / P(Document) C. P(Class|Document) = P(Class) - P(Document|Class) / P(Document) D. P(Class|Document) = P(Document|Class) * P(Class) / P(Document) 13. What is one of the core principles that underpins the functionality of MNB? A. Each document is represented as a single feature. B. Class probabilities are independent of the feature counts. C. Features are assumed to be represented as a vector of counts. D. Classification relies entirely on word order in documents.