Wayground logo

Free Printable Worksheets

Font size

S
M
L
XL
Worksheets

Machine Learning Quiz

Total questions: 20

Worksheet time: 10mins

Name
Class
Date
1.

What is precision in the context of machine learning metrics?

a)

The ratio of true positives to the total number of predictions.

b)

The ratio of true positives to the sum of true positives and false negatives.

c)

The ratio of true positives to the sum of true positives and false positives.

d)

The ratio of true negatives to the total number of predictions.

2.

Which of the following is an advantage of ensemble methods?

a)

They reduce the complexity of the models.

b)

They improve the robustness and accuracy of predictions.

c)

They require less data for training.

d)

They eliminate the need for cross-validation.

3.

In the voting ensemble method, the final prediction is made based on:

a)

The majority vote of the base models.

b)

The average of the predictions of the base models.

c)

The weighted sum of the predictions of the base models.

d)

All of the above.

4.

Bagging helps in improving model performance by:

a)

Reducing bias.

b)

Reducing variance.

c)

Increasing bias.

d)

Increasing variance.

5.

Which of the following is a boosting algorithm?

a)

Random Forest

b)

Gradient Boosting

c)

K-Nearest Neighbors

d)

Support Vector Machine

6.

In stacking, the model that combines the predictions of base models is called:

a)

Primary model

b)

Secondary model

c)

Meta model

d)

Base model

7.

A decision tree splits data at each node based on:

a)

Random selection

b)

The feature that gives the highest reduction in impurity

c)

The feature with the most missing values

d)

The feature with the least variance

8.

Which measure is commonly used to evaluate the splits in a decision tree for classification tasks?

a)

Variance

b)

Mean Absolute Error

c)

Gini Impurity

d)

Euclidean Distance

9.

In regression trees, the best split is chosen based on the reduction in:

a)

Gini Impurity

b)

Entropy

c)

Variance

d)

Accuracy

10.

Random Forests improve the accuracy of predictions by:

a)

Using a single decision tree

b)

Combining multiple decision trees trained on random subsets of the data

c)

Using linear regression

d)

Ignoring certain features

11.

Which metric is most appropriate to use when the dataset has a significant class imbalance?

a)

Accuracy

b)

Precision

c)

Recall

d)

F1-Score

12.

Which ensemble method involves training multiple models on different subsets of the training data with replacement?

a)

Voting

b)

Bagging

c)

Boosting

d)

Stacking

13.

In a weighted voting ensemble method, the model with the ________ accuracy typically has the highest weight.

a)

Lowest

b)

Highest

c)

Average

d)

Most recent

14.

Which of the following is a common algorithm used in bagging?

a)

Random Forest

b)

AdaBoost

c)

Gradient Boosting

d)

Support Vector Machines

15.

Boosting algorithms sequentially train models by giving more weight to ________.

a)

Correctly classified samples

b)

Misclassified samples

c)

Random samples

d)

Large samples

16.

The meta-model in stacking is trained on:

a)

The original dataset

b)

The predictions of the base models

c)

A subset of the features

d)

The residual errors of the base models

17.

In a decision tree, what is a leaf node?

a)

A node that splits into further sub-nodes

b)

A node that does not split further and represents a class label or value

c)

The topmost node in the tree

d)

A node with the highest impurity

18.

Which splitting criterion is used by default in the Sklearn implementation of decision trees for classification?

a)

Entropy

b)

Information Gain

c)

Gini Impurity

d)

Variance Reduction

19.

In the context of regression trees, which of the following measures is minimized to find the best split?

a)

Gini Impurity

b)

Mean Squared Error (MSE)

c)

Cross-Entropy

d)

Information Gain

20.

Random Forests create diversity among individual trees by:

a)

Using the entire dataset for each tree

b)

Using different subsets of features and data for each tree

c)

Training each tree sequentially

d)

Using the same features for all trees