Wayground logo

Free Printable Worksheets

Font size

S
M
L
XL
Worksheets

Understanding Convolutional Neural Networks

Total questions: 18

Worksheet time: 9mins

Name
Class
Date
1.

What is the principle of convolutional layers compared to fully-connected layers?

a)

Convolutional layers use a floating window to process pixels.

b)

Fully-connected layers are more efficient in computation.

c)

Convolutional layers are dependent on the shape of images.

d)

Fully-connected layers have fewer parameters.

2.

What is the objective of the convolution operation?

a)

To extract high-level features from the input image.

b)

To increase the number of parameters.

c)

To reduce the dimensions of the image.

d)

To apply filters to the image.

3.

What is the purpose of zero padding in convolutional layers?

a)

To preserve the size of the original image.

b)

To increase the number of parameters.

c)

To reduce the computational efficiency.

d)

To enhance the quality of the image.

4.

What does stride control in convolutional layers?

a)

The size of the filter.

b)

How the filter convolves through the input layer.

c)

The number of channels in the input layer.

d)

The dimensions of the output image.

5.

What is the role of pooling layers in CNNs?

a)

To reduce the dimensions of feature maps.

b)

To increase the number of parameters.

c)

To apply convolution operations.

d)

To flatten the output from convolutional layers.

6.

What is the main advantage of max pooling?

a)

It selects the maximum pixel value.

b)

It reduces the number of parameters.

c)

It increases the computational efficiency.

d)

It preserves all pixel values.

7.

What is the purpose of regression algorithms?

a)

To predict the value of an output variable.

b)

To classify data into categories.

c)

To reduce the dimensions of data.

d)

To enhance the quality of images.

8.

What does R Squared (R2) indicate in regression models?

a)

The performance of the model.

b)

The loss value of the model.

c)

The accuracy of predictions.

d)

The number of parameters in the model.

9.

Which layer in a CNN is primarily responsible for detecting edges and textures in an image?

a)

Pooling Layer

b)

Convolutional Layer

c)

Fully-Connected Layer

d)

Input Layer

10.

What is the role of padding in a convolutional layer?

a)

To increase the number of features

b)

To shrink the image size

c)

To preserve image dimensions after convolution

d)

To activate more neurons

11.

Which technique helps reduce the number of parameters in a CNN?

a)

Fully-connected layers

b)

Input resizing

c)

Pooling layers

d)

Adding more channels

12.

Why is Mean Absolute Error (MAE) used in regression model evaluation?

a)

It penalizes large errors more than small ones

b)

It shows percentage errors

c)

It gives the average magnitude of prediction errors

d)

It explains variance in data

13.

Which regression algorithm is best suited for modeling a linear relationship between variables?

a)

Decision Tree Regression

b)

Support Vector Regression

c)

Linear Regression

d)

Polynomial Regression

14.

A company observes that the sales data forms a curve when plotted against advertising spend. Which regression technique is most appropriate?

a)

Linear Regression

b)

Support Vector Regression

c)

Polynomial Regression

d)

Ridge Regression

15.

Which metric gives the average of the absolute errors between predicted and actual values?

a)

Mean Squared Error

b)

Mean Absolute Error

c)

Root Mean Squared Error

d)

R2 Squared

16.

Why is MSE sometimes preferred over MAE?

a)

It penalizes small errors more

b)

It penalizes larger errors more

c)

It is less sensitive to outliers

d)

It is always lower than MAE

17.

Which metric is best when you need to express error as a percentage?

a)

MAE

b)

RMSE

c)

MSE

d)

MAPE

18.

What does a high R² value (close to 1) indicate?

a)

High error

b)

Poor model performance

c)

Model explains most of the variability in target

d)

Model is underfitting