AI/ML Quiz - Common Jobs

AI/ML

1 / 40

1. How does a confusion matrix fit into the classification process?

1) Selecting hyperparameters

2) Visualizing decision boundaries

3) Handling missing data

4) Evaluating model performance

2 / 40

2. What part does the trade-off between exploration and exploitation play in reinforcement learning?

1) Balancing the trade-off between exploring new actions and exploiting known actions

2) All

3) Balancing the trade-off between precision and recall

4) Balancing the use of supervised and unsupervised learning

3 / 40

3. What is the unsupervised learning application of the Expectation-Maximization (EM) algorithm?

1) Maximizing the likelihood of the observed data

2) Iteratively estimating parameters for mixture models

3) Minimizing the reconstruction error in autoencoders

4) Imputing missing values in a dataset

4 / 40

4. In terms of propositional logic, which claim is untrue?

1) Predicate and Preposition

2) None

3) Predicate and Subject

4) Subject and an Object

5 / 40

5. Which algorithm is employed by the game tree to determine a win or loss?

1) DFS/BFS algorithm

2) Heuristic Search Algorithm

3) Greedy Search Algorithm

4) Min/Max algorithm

6 / 40

6. What does the term "class label" mean in classification?

1) The name of the model

2) The input features of a model

3) The output of a regression model

4) The predicted category of an input

7 / 40

7. Why is regularization used in machine learning models?

1) To make prejudice more pronounced

2) To reduce the complexity of the model

3) To expedite the training of models

4) In order to make the model more complex

8 / 40

8. In the context of decision trees, what does the term "entropy" mean?

1) The measure of impurity or disorder in a set of data

2) The depth of the decision tree

3) The number of leaf nodes in the tree

4) The ratio of training to testing data

9 / 40

9. In LISP programming, the square root is entered as_____.

1) Sqrt(x)

2) (sqrt x)

3) x/2

4) None

10 / 40

10. Why is feature scaling important in machine learning?

1) To standardize the range of features

2) To increase the complexity of models

3) To decrease the dimensionality of features

4) To remove outliers from the data

11 / 40

11. State statement is valid for the Heuristic function

1) The heuristic function does not have any return type.

2) The heuristic function calculates the cost of an optimal path between the pair of states.

3) The heuristic function is used to solve mathematical problems.

4) The heuristic function takes parameters of type string and returns an integer value.

12 / 40

12. What is supervised learning's main objective?

1) Learn from unlabeled data

2) Minimize errors in predictions

3) Maximize computational efficiency

4) Predict future events

13 / 40

13. Which of the subsequent describes a deterministic algorithm?

1) PCA

2) None

3) KNN

4) K-Means clustering

14 / 40

14. Among the languages listed, which one is not frequently used for AI?

1) PROLOG

2) Perl

3) LISP

4) Python

15 / 40

15. In the context of tree-based models such as Random Forests, what does feature importance mean?

1) The importance of including a feature in the dataset

2) The correlation between features

3) The contribution of each feature to the model's predictions

4) The number of times a feature appears in the dataset

16 / 40

16. The condition that an algorithm is considered complete is _______________.

1) It contains a loop

2) It does not end with a solution.

3) It begins with a solution.

4) It ends with a solution (if any exists).

17 / 40

17. How does the classification algorithm known as Support Vector Machine (SVM) operate?

1) It uses decision trees to make predictions

2) It finds the nearest neighbors for each data point

3) It fits a linear hyperplane to separate classes

4) It performs clustering based on data density

18 / 40

18. When a hypothesis predicts a good outcome but the actual result is negative, this situation is referred to as _______.

1) False Positive Hypothesis

2) False Negative Hypothesis

3) Specialized Hypothesis

4) Consistent Hypothesis

19 / 40

19. Why is feature engineering used in machine learning?

1) Building better hardware

2) Extracting features from images

3) Selecting the most relevant features

4) Engineering new features using deep learning

20 / 40

20. What effect does the loss function selection have on a machine learning model's training process?

1) The choice of loss function only impacts model evaluation

2) The loss function defines the model's architecture

3) The loss function determines the optimization objective

4) The loss function has no impact on training

21 / 40

21. Who is credited with creating artificial intelligence?

1) Charles Babbage

2) Alan Turing

3) Andrew Ng

4) John McCarthy

22 / 40

22. Which of the following statements about conditional probability is true?

1) Conditional Probability has no effect or relevance on independent events.

2) None

3) Conditional Probability can be applied to a single event.

4) Conditional Probability gives 100% accurate results.

23 / 40

23. What distinguishes recall from precision?

1) Precision focuses on false negatives, recall focuses on false positives

2) Precision and recall are unrelated metrics

3) Both measure the same thing

4) Precision focuses on false positives, recall focuses on false negatives

24 / 40

24. The deepest point at which alpha-beta pruning can be implemented

1) Eight states

2) Six states

3) Ten states

4) Any depth

25 / 40

25. Describe the meaning of "latent variables" in graphical models of probability.

1) Variables used to encode temporal information

2) Variables that are directly measured in the dataset

3) Variables representing missing data

4) Variables that are not observed directly but inferred from observed variables

26 / 40

26. What is the difference between bagging and boosting in ensemble learning?

1) Bagging trains each model independently, boosting focuses on examples misclassified by previous models

2) Bagging combines predictions using voting, boosting combines predictions using weighted averaging

3) Bagging increases model diversity, boosting decreases it

4) Bagging trains models sequentially, boosting trains them in parallel

27 / 40

27. When it comes to machine learning models, what does overfitting mean?

1) Fitting a model to the validation set

2) fitting a model with insufficient data

3) Fitting a model with too few features

4) Fitting a model too closely to the training data

28 / 40

28. What function does the learning_rate parameter serve in the optimization of gradient descent?

1) The degree of regularization used in the model

2) The amount of work done in each cycle

3) The quantity of optimization process iterations

4) The rate of convergence of the algorithm

29 / 40

29. How does the model's performance be affected by the bias-variance trade-off?

1) Better models have low variance and low bias.

2) High variance can result in overfitting, whereas high bias can cause underfitting.

3) Raising bias raises the accuracy of the model.

4) A higher variance enhances the generalization of the model.

30 / 40

30. Which of the following scenarios calls for the use of blind search?

1) None

2) Real-life Simulation

3) Advanced Game Theory

4) Small Search Space

31 / 40

31. In LISP, the addition of 5+8 is entered as_______.

1) 5+8=

2) 5 add 8

3) 5+8

4) (+5 8)

32 / 40

32. What is the purpose of the activation function in a neural network?

1) Defines the learning rate

2) Introduces non-linearity to the model

3) Sets the number of neurons in each layer

4) Controls the initialization of weights

33 / 40

33. A method created to ascertain whether a machine was capable of exhibiting the artificial intelligence known as the ___

1) Boolean Algebra

2) Turing Test

3) Algorithm

4) Logarithm

34 / 40

34. Knowledge in artificial intelligence can be expressed as _______.

i.Predicate Logic

ii.Propositional Logic

iii. Compound Logic

iv.Machine Logic

1) Only II

2) Both II and III

3) Only IV

4) Both I and II

35 / 40

35. What does the k-nearest neighbors (KNN) algorithm aim to achieve?

1) Clustering data into k groups

2) Classifying data based on its neighbors

3) Reducing the dimensionality of features

4) Predicting a continuous output

36 / 40

36. What is the purpose of cross-validation in machine learning?

1) Cross-training different models

2) Testing a model's generalization ability

3) Evaluating model performance on multiple datasets

4) Selecting hyperparameters

37 / 40

37. What effect does a kernel selection have on a Support Vector Machine's (SVM) performance?

1) The kernel influences the learning rate during training

2) The kernel determines the SVM's maximum margin

3) The kernel has no impact on SVM performance

4) The kernel defines the transformation of input features into a higher-dimensional space

38 / 40

38. Which of the following describes an algorithm for unsupervised learning?

1) Decision Trees

2) Linear Regression

3) Support Vector Machines

4) K-Means Clustering

39 / 40

39. What does dropout in neural networks mean?

1) Reduces the model's learning rate

2) Regularizes the model to prevent overfitting

3) Ensures convergence during training

4) Increases the number of neurons in each layer

40 / 40

40. What does the neural network training epochs parameter mean?

1) The learning rate for weight updates

2) The number of complete passes through the entire training dataset

3) The number of layers in the neural network

4) The number of training examples processed in one iteration

Your score is

The average score is 37%

About Us

Company

Education

Quick Links

Get latest job updates

Connect With Us