Free Practice Quiz Question List

3 Training and Testing Models Online Quiz Questions

Use this free practice quiz with 20 questions to review 3 Training and Testing Models, test your knowledge, and prepare for your next test or exam.

20 questions

Choose one

1 point

During supervised training, a model produces predictions for examples whose targets are known. What is the loss function's role?

A
It adjusts the model's parameters to reduce prediction error.
B
It measures how far predictions are from known targets.
C
It applies learned parameters to new input data.
D
It guarantees that each prediction is correct.

True or false

1 point

A trained home-price model receives the size of a new home and produces a price estimate. This is inference, and it does not ordinarily update the model's parameters.

A
True
B
False

Choose one

1 point

What do we call the output a model produces for a particular input?

A
A target
B
A parameter
C
A prediction
D
A loss function

Choose all

1 point

A supervised-learning dataset is being prepared to predict home sale prices. Select all statements that correctly describe the examples.

A
A home's characteristics can serve as input features.
B
A home's characteristics are the target sale price.
C
The known sale price can serve as the target.
D
A supervised example must omit its target.

True or false

1 point

A training process may repeat over examples or batches for multiple passes, stopping according to a chosen rule.

A
True
B
False

Fill in the blank

1 point

During training, a measures how far predictions are from targets, while an adjusts parameters to reduce that measure.

Choose one

1 point

In scikit-learn, an estimator has already been fitted. Which operation should be used to obtain its prediction for a new input without training it again?

A
Use fit on every new input to obtain a prediction.
B
Use predict to change the estimator's parameters.
C
Use fit to measure the prediction for a new input.
D
Use predict with the fitted estimator to obtain an output.

Written response

1 point

A home-price model uses the formula $\text{predicted price} = \text{weight} \times \text{size} + \text{bias}$ . If the weight is $2$ , the size is $3$ , and the bias is $5$ , what numerical value does the model predict?

Choose all

1 point

Which actions are part of a typical model-training process? Select all that apply.

A
Produce predictions for training examples.
B
Measure the predictions against known targets.
C
Apply learned parameters to new inputs without changing them.
D
Adjust parameters to reduce the measured loss.

Choose one

1 point

What makes up a trained model after its parameter values have been learned?

A
Only the examples used during training
B
The learned parameter values together with the model's structure
C
Only the loss function and optimization method
D
The predictions made for the training examples

Written response

1 point

Enter the single term for the output a model produces for an input.

Open ended

1 point

An email classifier produces a spam score and then assigns the email a “spam” or “not spam” label. Explain how the score and label relate, and whether the model's result is guaranteed to be correct.

True or false

1 point

A machine-learning model's prediction is always guaranteed to be correct.

A
True
B
False

Choose one

1 point

A supervised-learning example contains a home’s characteristics and its known sale price. What role does the known sale price play?

A
It is an input feature used to describe the home.
B
It is the target the model learns to predict from the home’s features.
C
It is a parameter that the model adjusts during training.
D
It is the loss function used to measure prediction error.

Choose one

1 point

During training, a model’s predictions differ from known targets. Which description correctly distinguishes the roles of the loss function and the optimization method?

A
The optimization method measures prediction error, and the loss function changes the parameters.
B
The loss function produces predictions, and the optimization method supplies the known targets.
C
The loss function measures how far predictions are from targets, and the optimization method adjusts parameters to reduce that difference.
D
The loss function selects new input features, and the optimization method decides the model’s output type.

Choose one

1 point

A fitted home-price model receives the size of a new home and returns a price estimate. The model’s parameters are not changed. What is the model doing?

A
It is training because every new input automatically changes the model’s parameters.
B
It is inference because the trained model applies its learned parameters to produce an output.
C
It is training because the model must first compare the new input with a known target.
D
It is neither training nor inference because the input has not yet been assigned a class.

Written response

1 point

In the model expression $\text{predicted price} = \text{weight} \times \text{size} + \text{bias}$ , what is the name of the parameter that multiplies size?

Written response

1 point

A home-price model uses $\text{predicted price} = \text{weight} \times \text{size} + \text{bias}$ . For a home with size $6$ , weight $2.5$ , and bias $4$ , what value does the model predict?

Choose one

1 point

In scikit-learn, which pairing correctly describes the operations used to train an estimator and then obtain its predictions?

A
fit trains the estimator; predict obtains outputs from the fitted estimator.
B
predict trains the estimator; fit obtains outputs from it.
C
Both fit and predict train the estimator, but on different inputs.
D
fit measures the loss; predict adjusts the parameters to reduce it.

Fill in the blank

1 point

A home-price model receives the size of a new home and predicts its price. The sale price is not yet known, and the input is used only to make a prediction. It is not yet a .