Machine Learning MCQs Quiz Multiple Choice Questions & Answers

1. Regression trees are often used to model ........... data.

1. linear
2. nonlinear
3. categorical
4. symmetrical

2. Selecting data so as to assure that each class is properly represented in both the training and test set.

1. cross validation
2. stratification
3. verification
4. bootstrapping

3. Simple regression assumes a ........... relationship between the input attribute and output attribute.

1. linear
2. quadratic
3. reciprocal
4. inverse

4. Supervised learning and unsupervised clustering both require at least one

1. hidden attribute.
2. output attribute.
3. input attribute.
4. categorical attribute.

5. Supervised learning differs from unsupervised clustering in that supervised learning requires

1. at least one input attribute.
2. input attributes to be categorical.
3. at least one output attribute.
4. ouput attriubutes to be categorical.

6. Suppose your model is overfitting. Which of the following is NOT a valid way to try and reduce the overfitting?

1. Increase the amount of training data.
2. Improve the optimisation algorithm being used for error minimisation.
3. Decrease the model complexity.
4. Reduce the noise in the training data.

7. The adjusted multiple coefficient of determination accounts for

1. the number of dependent variables in the model
2. the number of independent variables in the model
3. unusually large predictors
4. none of the above

8. The average positive difference between computed and desired outcome values.

1. root mean squared error
2. mean squared error
3. mean absolute error
4. mean positive error

9. The average squared difference between classifier predicted output and actual output.

1. mean squared error
2. root mean squared error
3. mean absolute error
4. mean relative error

10. The correlation between the number of years an employee has worked for a company and the salary of the employee is 0.75. What can be said about employee salary and years worked?

1. There is no relationship between salary and years worked.
2. Individuals that have worked for the company the longest have higher salaries.
3. Individuals that have worked for the company the longest have lower salaries.
4. The majority of employees have been with the company a long time.

11. The correlation coefficient for two real-valued attributes is 0.85. What does this value tell you?

1. The attributes are not linearly related.
2. As the value of one attribute increases the value of the second attribute also increases.
3. As the value of one attribute decreases the value of the second attribute increases.
4. The attributes show a curvilinear relationship.

12. The leaf nodes of a model tree are

1. averages of numeric output attribute values.
2. nonlinear regression equations.
3. linear regression equations.
4. sums of numeric output attribute values.

13. The multiple coefficient of determination is computed by

1. dividing SSR by SST
2. dividing SST by SSR
3. dividing SST by SSE
4. none of the above

14. The process of forming general concept definitions from examples of concepts to be learned.

1. Deduction
2. abduction
3. induction
4. conjunction

15. The standard error is defined as the square root of this computation.

1. The sample variance divided by the total number of sample instances.
2. The population variance divided by the total number of sample instances.
3. The sample variance divided by the sample mean.
4. The population variance divided by the sample mean.

16. This clustering algorithm initially assumes that each data instance represents a single cluster.

1. agglomerative clustering
2. conceptual clustering
3. K-Means clustering
4. expectation maximization

17. This clustering algorithm merges and splits nodes to help modify nonoptimal partitions.

1. agglomerative clustering
2. expectation maximization
3. conceptual clustering
4. K-Means clustering

18. This supervised learning technique can process both numeric and categorical input attributes.

1. linear regression
2. Bayes classifier
3. logistic regression
4. backpropagation learning

19. This technique associates a conditional probability value with each data instance.

1. linear regression
2. logistic regression
3. simple regression
4. multiple linear regression

20. This unsupervised clustering algorithm terminates when mean values computed for the current iteration of the algorithm are identical to the computed mean values for the previous iteration.

1. agglomerative clustering
2. conceptual clustering
3. K-Means clustering
4. expectation maximization

21. When doing least-squares regression with regularisation (assuming that the optimisation can be done exactly), increasing the value of the regularisation parameter (Lambda)

1. will never decrease the training error.
2. will never increase the training error.
3. will never decrease the testing error.
4. will never increase the testing error.

22. Which is not true about Gradient of a continuous and differentiable function

1. is zero at a minimum
2. is non-zero at a maximum
3. is zero at a saddle point
4. decreases as you get closer to the minimum

23. Which of the following is a common use of unsupervised clustering?

1. detect outliers
2. determine a best set of input attributes for supervised learning
3. evaluate the likely performance of a supervised learner model
4. determine if meaningful relationships can be found in a dataset

24. Which of the following is not an advantage of Grid search

1. It can be applied to non-differentiable functions.
2. It can be applied to non-continuous functions.
3. It is easy to implement.
4. It runs reasonably fast for multiple linear regression.

25. Which of the following points would Bayesians and frequentists disagree on?

1. The use of a non-Gaussian noise model in probabilistic regression.
2. The use of probabilistic modelling for regression.
3. The use of prior distributions on the parameters in a probabilistic model.
4. The use of class priors in Gaussian Discriminant Analysis

26. Which of the following sentence is FALSE regarding regression?

1. It relates inputs to outputs.
2. It is used for prediction.
3. It may be used for interpretation.
4. It discovers causal relationships.

27. Which statement about outliers is true?

1. Outliers should be identified and removed from a dataset.
2. Outliers should be part of the training dataset but should not be present in the test data.
3. Outliers should be part of the test dataset but should not be present in the training data.
4. The nature of the problem determines how outliers are used.

28. Which statement is true about neural network and linear regression models?

1. Both models require input attributes to be numeric.
2. Both models require numeric attributes to range between 0 and 1.
3. The output of both models is a categorical attribute value.
4. Both techniques build models whose output is determined by a linear sum of weighted input attribute values.

29. Which statement is true about prediction problems?

1. The output attribute must be categorical.
2. The output attribute must be numeric.
3. The resultant model is designed to determine future outcomes.
4. The resultant model is designed to classify current behavior.

30. With Bayes classifier, missing data items are

1. treated as equal compares.
2. treated as unequal compares.
3. replaced with a default value.
4. ignored.

31. You observe the following while fitting a linear regression to the data: As you increase the amount of training data, the test error decreases and the training error increases. The train error is quite low (almost what you expect it to), while the test error is much higher than the train error. What do you think is the main reason behind this behavior. Choose the most probable option.

1. High variance
2. High model bias
3. High estimation bias
4. None of the above

32. A measure of goodness of fit for the estimated regression equation is the

1. multiple coefficient of determination
2. mean square due to error
3. mean square due to regression
4. none of the above

33. A nearest neighbor approach is best used

1. with large-sized datasets.
2. when irrelevant attributes have been removed from the data.
3. when a generalized model of the data is desireable.
4. when an explanation of what has been found is of primary importance.

34. A regression model in which more than one independent variable is used to predict the dependent variable is called

1. a simple linear regression model
2. a multiple regression models
3. an independent model
4. none of the above

35. A term used to describe the case when the independent variables in a multiple regression model are correlated is

1. regression
2. correlation
3. multicollinearity
4. none of the above

36. Adding more basis functions in a linear model... (pick the most probably option)

1. Decreases model bias
2. Decreases estimation bias
3. Decreases variance
4. Doesnt affect bias and variance

37. Another name for an output attribute.

1. predictive variable
2. independent variable
3. estimated variable
4. dependent variable

38. Bootstrapping allows us to

1. choose the same training instance several times.
2. choose the same test set instance several times.
3. build models with alternative subsets of the training data several times.
4. test a model with alternative subsets of the test data several times.

39. Choose the options that are correct regarding machine learning (ML) and artificial intelligence (AI),(A) ML is an alternate way of programming intelligent machines.(B) ML and AI have very different goals.(C) ML is a set of techniques that turns a dataset into a software.(D) AI is a software that can emulate the human mind.

1. (A), (B), (D)
2. (A), (C), (D)
3. (B), (C), (D)
4. All are correct

40. Classification problems are distinguished from estimation problems in that

1. classification problems require the output attribute to be numeric.
2. classification problems require the output attribute to be categorical.
3. classification problems do not allow an output attribute.
4. classification problems are designed to predict future outcome.

41. Computational complexity of Gradient descent is

1. linear in D
2. linear in N
3. polynomial in D
4. dependent on the number of iterations

42. Computers are best at learning

1. facts.
2. concepts.
3. procedures.
4. principles.

43. Consider a binary classification problem. Suppose I have trained a model on a linearly separable training set, and now I get a new labeled data point which is correctly classified by the model, and far away from the decision boundary. If I now add this new point to my earlier training set and re-train, in which cases is the learnt decision boundary likely to change?

1. When my model is a perceptron and logistic regression.
2. When my model is logistic regression and Gaussian discriminant analysis.
3. When my model is an SVM.
4. When my model is a perceptron

44. Data used to build a data mining model.

1. validation data
2. training data
3. test data
4. hidden data

45. Data used to optimize the parameter settings of a supervised learner model.

1. training
2. test
3. verification
4. validation

46. Generalization error measures how well an algorithm perform on unseen data. The test error obtained using cross-validation is an estimate of the generalization error. Is this estimate unbiased?

1. Yes
2. No

47. Grid search is

1. Linear in D and Polynomial in D
2. Polynomial in D
3. Exponential in D and Linear in N
4. Polynomial in D and Linear in N

48. K-fold cross-validation is

1. linear in K
2. quadratic in K
3. cubic in K
4. exponential in K

49. Let us say that we have computed the gradient of our cost function and stored it in a vector g. What is the cost of one gradient descent update given the gradient?

1. O(D)
2. O(N)
3. O(ND)
4. O(ND²)

50. Logistic regression is a ........... regression technique that is used to model data having a ........... outcome.

1. linear, numeric
2. linear, binary
3. nonlinear, numeric
4. nonlinear, binary