01
Choose one
1 point
A neural network is being used to estimate tomorrow’s demand as a number. Which part produces that estimate?
Free Practice Quiz Question List
5 Neural Networks Online Quiz Questions
Use this free practice quiz with 20 questions to review 5 Neural Networks, test your knowledge, and prepare for your next test or exam.
20 questions
02
True or false
1 point
True or false: In a basic neural network, the input layer receives features from an example, such as pixel values or table measurements.
- A
True
- B
False
03
Written response
1 point
What is the two-word name for the process in which an input passes through the network layer by layer to produce a prediction?
04
Choose one
1 point
After a network makes a prediction, which tool compares it with the correct answer to measure error?
- A
A loss function
- B
An activation function
- C
A validation set
- D
An input layer
05
True or false
1 point
True or false: Backpropagation calculates how changing each parameter would affect the loss, while an optimization method such as gradient descent then adjusts the parameters.
- A
True
- B
False
06
Choose all
1 point
Which tasks are identified as potential uses of neural networks? Select all that apply.
- A
Recognizing objects in images
- B
Translating text
- C
Guaranteeing predictions without training examples
- D
Estimating demand
07
Written response
1 point
Which two-word optimization method nudges parameters in directions expected to reduce the loss?
08
Choose one
1 point
Why does a neural network use nonlinear activation functions between computations?
- A
They guarantee a correct prediction.
- B
They prevent the network from needing training examples.
- C
They allow the network to represent nonlinear relationships.
- D
They make every network’s predictions easy to interpret.
09
Choose all
1 point
Which are limitations or cautions associated with neural networks? Select all that apply.
- A
They may require substantial data and computing resources.
- B
They may learn unwanted patterns from training data.
- C
They can be difficult to interpret.
- D
They are automatically the best choice for every problem.
10
Open ended
1 point
Explain the main steps in the neural network training cycle, including how prediction error leads to parameter updates and what is done after an update.
11
Choose one
1 point
Which pair names the learned numerical parameters of a neural network?
- A
Features and predictions
- B
Weights and biases
- C
Input and output layers
- D
Training and validation examples
12
Choose one
1 point
A model receives pixel values from an image before processing them. Which neural network component receives those features?
- A
The hidden layer transforms the features.
- B
The input layer receives the example's features.
- C
The output layer produces the prediction.
- D
The activation function compares the prediction with the correct answer.
13
True or false
1 point
True or false: If a neural network stacks layers but uses no nonlinear activation functions, it can still represent nonlinear relationships.
- A
True
- B
False
14
Written response
1 point
What is the term for the nonlinear transformation applied to a neural network unit’s result?
15
Choose one
1 point
After a neural network makes a prediction, which component compares it with the correct answer to measure the error?
- A
A loss function
- B
An activation function
- C
A validation set
- D
Backpropagation
16
Written response
1 point
What single term refers collectively to a neural network's learned weights and biases?
17
Choose one
1 point
A hidden unit receives values from the preceding layer. Which description best matches what it does?
- A
A hidden unit combines incoming values using weights and a bias, then applies an activation function.
- B
A hidden unit compares the final prediction with the correct answer.
- C
A hidden unit checks whether the model generalizes to new examples.
- D
A hidden unit selects training examples for the optimization method.
18
Choose one
1 point
During training, a model uses backpropagation and gradient descent. Which statement correctly distinguishes their roles?
- A
Backpropagation updates parameters, while gradient descent calculates the prediction.
- B
Backpropagation calculates how parameter changes affect loss, while gradient descent nudges parameters to reduce it.
- C
Backpropagation checks generalization, while gradient descent measures the prediction error.
- D
Backpropagation transforms inputs, while gradient descent applies the activation function.
19
Fill in the blank
1 point
A team wants to check whether a model trained on one group of examples can predict well on separate examples. It reserves a for this check.
20
Fill in the blank
1 point
A network performs well on its training examples but poorly on a separate validation set. This is evidence of .