Loss functions (MSE, cross-entropy) in TensorFlow - Interactive Code Practice

Practice - 5 Tasks

1fill in blank

easy

Complete the code to import the mean squared error loss from TensorFlow.

TensorFlow

from tensorflow.keras.losses import [1]

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

AMeanSquaredError

Bmean_squared_error

CMSE

DCrossEntropy

Attempts:

3 left

💡 Hint

Common Mistakes

Using function names instead of class names for import.

Confusing cross-entropy loss with mean squared error.

The correct import for mean squared error loss function in TensorFlow Keras is MeanSquaredError class.

2fill in blank

medium

Complete the code to compile a model using mean squared error loss.

TensorFlow

model.compile(optimizer='adam', loss=[1], metrics=['accuracy'])

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

A'mean_squared_error'

B'hinge'

C'cross_entropy'

D'categorical_crossentropy'

Attempts:

3 left

💡 Hint

Common Mistakes

Using cross-entropy loss names when mean squared error is needed.

Forgetting to put the loss name in quotes.

To use mean squared error as the loss function, specify 'mean_squared_error' as the loss parameter.

3fill in blank

hard

Fix the error in the code to use sparse categorical cross-entropy loss correctly.

TensorFlow

model.compile(optimizer='adam', loss=[1], metrics=['accuracy'])

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

A'mean_squared_error'

B'sparse_categorical_crossentropy'

C'binary_crossentropy'

D'categorical_crossentropy'

Attempts:

3 left

💡 Hint

Common Mistakes

Using categorical cross-entropy with sparse labels.

Using mean squared error for classification tasks.

The correct loss function for sparse categorical labels is 'sparse_categorical_crossentropy'.

4fill in blank

hard

Fill both blanks to create a model with binary cross-entropy loss and sigmoid activation.

TensorFlow

model = tf.keras.Sequential([
    tf.keras.layers.Dense(1, activation=[1])
])
model.compile(optimizer='adam', loss=[2], metrics=['accuracy'])

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

A'sigmoid'

B'relu'

C'binary_crossentropy'

D'mean_squared_error'

Attempts:

3 left

💡 Hint

Common Mistakes

Using relu activation for output layer in binary classification.

Using mean squared error loss for classification.

Binary classification models use sigmoid activation and binary cross-entropy loss.

5fill in blank

hard

Fill all three blanks to define a custom loss function using mean squared error.

TensorFlow

import tensorflow as tf

def custom_loss(y_true, y_pred):
    return tf.keras.losses.[1](y_true, y_pred)

model.compile(optimizer='adam', loss=[2], metrics=[[3]])

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

Amean_squared_error

B'mean_squared_error'

C'accuracy'

D'categorical_accuracy'

Attempts:

3 left

💡 Hint

Common Mistakes

Putting quotes around the function call inside custom loss.

Using wrong metric names.

The custom loss calls mean_squared_error function. The compile loss is the string name. Accuracy is a common metric.