Prediction and evaluation in TensorFlow - Interactive Code Practice

Practice - 5 Tasks

1fill in blank

easy

Complete the code to make predictions using the trained model.

TensorFlow

predictions = model.[1](x_test)

Drag options to blanks, or click blank then click option'

Afit

Bcompile

Cevaluate

Dpredict

Attempts:

3 left

💡 Hint

Common Mistakes

Using 'fit' instead of 'predict' which trains the model.

Using 'evaluate' which returns loss and metrics, not predictions.

The predict method is used to get the model's predictions on new data.

2fill in blank

medium

Complete the code to evaluate the model on test data and get the loss.

TensorFlow

loss, accuracy = model.[1](x_test, y_test)

Drag options to blanks, or click blank then click option'

Aevaluate

Bcompile

Cfit

Dpredict

Attempts:

3 left

💡 Hint

Common Mistakes

Using 'predict' which returns predictions, not loss or accuracy.

Using 'fit' which trains the model instead of evaluating.

The evaluate method returns the loss and metrics on the test data.

3fill in blank

hard

Fix the error in the code to correctly compute accuracy from predictions.

TensorFlow

accuracy = tf.keras.metrics.Accuracy()
accuracy.update_state(y_true, [1])
result = accuracy.result().numpy()

Drag options to blanks, or click blank then click option'

Ay_pred

Bmodel.predict(x_test)

Cy_test

Dy_true

Attempts:

3 left

💡 Hint

Common Mistakes

Passing true labels twice instead of predictions.

Passing the model call instead of prediction results.

The update_state method needs the predicted labels, which are y_pred.

4fill in blank

hard

Fill both blanks to compute predictions and then evaluate loss and accuracy.

TensorFlow

y_pred = model.[1](x_val)
loss, acc = model.[2](x_val, y_val)

Drag options to blanks, or click blank then click option'

Apredict

Bevaluate

Cfit

Dcompile

Attempts:

3 left

💡 Hint

Common Mistakes

Using 'fit' instead of 'evaluate' for performance metrics.

Using 'compile' which only sets up the model.

First, use predict to get predictions, then evaluate to get loss and accuracy.

5fill in blank

hard

Fill all three blanks to compute accuracy using TensorFlow metrics after prediction.

TensorFlow

y_pred = model.[1](x_test)
accuracy_metric = tf.keras.metrics.[2]()
accuracy_metric.update_state(y_test, [3])
accuracy = accuracy_metric.result().numpy()

Drag options to blanks, or click blank then click option'

Apredict

BAccuracy

Cy_pred

DPrecision

Attempts:

3 left

💡 Hint

Common Mistakes

Using 'Precision' metric instead of 'Accuracy'.

Passing true labels twice instead of predicted labels.

Use predict to get predictions, Accuracy metric to compute accuracy, and update state with predicted labels y_pred.