Machine Learning MCQs-2 (Performance Metrics, Linear Regression, Naïve Bayes Classifier )

                       Machine Learning MCQs- 2
Performance Metrics, Linear Regression, Naïve Bayes Classifier 

1.  The greater the value for ROC AUC, better the model:

1. True
2. False

Ans: 1
 
2.  A set of data are all close to each other, and they are close to the actual value.  This set of data can be described as...

1. Accurate
2. Precise
3. both Precise and accurate
4. None of the above

Ans: 3

3. The maximum value of the ROC AUC is:

1. 0.8
2. 0.9
3. 1
4. 0

Ans: 3

4. Recall can be increased by increasing the decision threshold. True or False?

1. True
2. False

Ans: 2

5. Which of these is a good measure to decide which threshold to use?

1. Confusion matrix
2. F1 score
3. ROC curve
4. Precision & Recall versus Threshold Curve

Ans: 4

6. Which of these may have to be performed before analyzing and training the dataset?

1. Shuffling
2. Cross-Validation
3. F1 Score
4. None

Ans: 1

7. For the below confusion matrix, what is the total number of training datasets?

	Not 5	5
Not 5	53272	1307
5	1077	4344

1. 50000
2. 60000
3. 70000
4. 80000

Ans: 2

8. For the below confusion matrix, what is the accuracy?

	Not 5	5
Not 5	53272	1307
5	1077	4344

1. 95%
2. 90%
3. 96%
4. 98%

Ans: 4

9.  For the below confusion matrix, what is the recall?

	Not 5	5
Not 5	53272	1307
5	1077	4344

1. 0.7
2. 0.8
3. 0.9
4. 0.95

Ans: 2

10. For the below confusion matrix, what is the precision?

	Not 5	5
Not 5	53272	1307
5	1077	4344

1. 0.73
2. 0.76
3. 0.78
4. 0.82

Ans: 2

11. F1 score is:

1. absolute mean of precision and recall
2. harmonic mean of precision and recall
3. squared mean of precision and recall
4. None

Ans: 2

12. For the below confusion matrix, what is the F1 score?

	Not 5	5
Not 5	53272	1307
5	1077	4344

1. 0.72
2. 0.784
3. 0.82
4. 0.84

Ans: 2

13. For a model to detect videos that are unsafe for kids, we need (safe video = positive class)

1. High precision, low recall
2. High recall, low precision
3. High Precision, High Recall
4. Low Precision, Low Recall

Ans: 1

14. For a model to detect shoplifters in surveillance images, we need (shoplifter is positive class)

1. High precision, low recall
2. High recall, low precision
3. High Precision, High Recall
4. Low Precision, Low Recall

Ans: 2 

15. which of these provide out-of-core support for linear regression?

1. Normal Equation
2. SGD
3. Batch Gradient Descent
4. None

Ans: 2

16. NormalEquation class in scikit-learn solve linear regression using:

1. Normal Equation
2. SGD
3. There is no NormalEquation class in scikit-learn
4. Mini-Batch Gradient Descent

Ans: 3

17.Training Linear Regression model using Normal Equation is linear with both the number of training dataset that have to be made and the number of features:

1. True
2. False

Ans: 2

18.Prediction using Linear Regression model is linear with both the number of predictions that have to be made and the number of features:

1. True
2. False

Ans: 1

19.Which of these is more prone to overfitting?

1. Linear Regression
2. Polynomial Regression

Ans: 2

20. Naive Baye is?

1. Conditional Independence
2. Conditional Dependence
3. Both 1 and 2
4. None of the above

Ans:1

21. Naive Bayes requires?

1. Categorical Values
2. Numerical Values
3. Either 1 or 2
4. Both 1 and 2

Ans: 1

22. Probabilistic Model of data within each class is?

1. Discriminative classification
2. Generative classification
3. Probabilistic classification
4. Both 2 and 3

Ans: 4

23. A Classification rule is said?

1. Discriminative classification
2. Generative classification
3. Probabilistic classification
4. Both 1 and 3

Ans: 4

24. Spam Classification is an example for ?

1. Naive Bayes
2. Probabilistic condition
3. Random Forest
4. All the Above

Ans: 1.

25. Naïve Bayes Algorithm is a ________ learning algorithm.

1. Supervised
2. Reinforcement
3. Unsupervised
4. None of these

Ans: 1

26. Which of the following is correct about the Naive Bayes?

1. Assumes that all the features in a dataset are independent
2. Assumes that all the features in a dataset are equally important
3. Both 1 and 2
4. None of the above

Ans: 2

27. Types of Naïve Bayes Model:

1. Gaussian
2. Multinomial
3. Bernoulli
4. All of the above

Ans: 4

28. Disadvantages of Naïve Bayes Classifier:

1. Naïve Bayes assumes that all features are independent or unrelated, so it cannot learn the relationship between features.
2. It performs well in Multi-class predictions as compared to the other Algorithms.
3. Naïve Bayes is one of the fast and easy ML algorithms to predict a class of datasets
4. It is the most popular choice for text classification problems

Ans: 1

29. The benefit of Naïve Bayes:-

1. Naïve Bayes is one of the fast and easy ML algorithms to predict a class of datasets.
2. It is the most popular choice for text classification problems.
3. It can be used for Binary as well as Multi-class Classifications.
4. All of the above

Ans: 4

30. TPR =

1. TP/(FN+TP)
2. TN/(TN+FP)
3. FP/(FP+TN)
4. FN/(FN+TP)

Ans: 1

31. TNR=

1. TP/(FN+TP)
2. TN/(TN+FP)
3. FP/(FP+TN)
4. FN/(FN+TP)

Ans: 2

32. FPR=

1. TP/(FN+TP)
2. TN/(TN+FP)
3. FP/(FP+TN)
4. FN/(FN+TP)

Ans: 3

33. FNR=

1. TP/(FN+TP)
2. TN/(TN+FP)
3. FP/(FP+TN)
4. FN/(FN+TP)

Ans: 4

34. Precision =

1. TP/(TP+FP)
2. TP/(FN+TP)
3. FP/(TP+FP)
4. FP/(FN+TP)

Ans: 1

35. Recall=

1. TP/(TP+FP)
2. TP/(FN+TP)
3. FP/(TP+FP)
4. FP/(FN+TP)

Ans: 2