Multi-class classification model in TensorFlow - Interactive Code Practice

Practice - 5 Tasks

1fill in blank

easy

Complete the code to create a multi-class classification model using TensorFlow Keras.

TensorFlow

model = tf.keras.Sequential([
    tf.keras.layers.Dense(64, activation='relu', input_shape=(input_shape,)),
    tf.keras.layers.Dense(10, activation=[1])
])

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

A'sigmoid'

B'linear'

C'softmax'

D'tanh'

Attempts:

3 left

💡 Hint

Common Mistakes

Using 'sigmoid' activation which is for binary classification.

Using 'linear' activation which does not output probabilities.

For multi-class classification, the output layer uses the softmax activation to output probabilities for each class.

2fill in blank

medium

Complete the code to compile the multi-class classification model with the correct loss function.

TensorFlow

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

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

A'categorical_crossentropy'

B'binary_crossentropy'

C'mean_squared_error'

D'hinge'

Attempts:

3 left

💡 Hint

Common Mistakes

Using binary_crossentropy which is for binary classification.

Using mean_squared_error which is for regression tasks.

For multi-class classification with one-hot encoded labels, categorical_crossentropy is the correct loss function.

3fill in blank

hard

Fix the error in the code to correctly prepare labels for multi-class classification training.

TensorFlow

y_train_encoded = tf.keras.utils.to_categorical([1], num_classes=10)

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

AX_train

BX_test

Cy_test

Dy_train

Attempts:

3 left

💡 Hint

Common Mistakes

Encoding input features instead of labels.

Encoding test labels instead of training labels.

Labels for training (y_train) must be converted to one-hot encoding for multi-class classification.

4fill in blank

hard

Fill both blanks to train the model with the encoded labels and set the number of epochs.

TensorFlow

history = model.fit(X_train, [1], epochs=[2], batch_size=32, validation_split=0.2)

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

Ay_train_encoded

By_test_encoded

C10

D5

Attempts:

3 left

💡 Hint

Common Mistakes

Using test labels for training.

Setting epochs too low or too high without reason.

Use the encoded training labels y_train_encoded and set epochs to 10 for training.

5fill in blank

hard

Fill all three blanks to evaluate the model on test data and print the accuracy.

TensorFlow

loss, accuracy = model.evaluate([1], [2], batch_size=32)
print(f"Test accuracy: [3]")

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

AX_test

By_test_encoded

Caccuracy

Dloss

Attempts:

3 left

💡 Hint

Common Mistakes

Using training data for evaluation.

Printing loss instead of accuracy.

Evaluate the model on test features X_test and encoded test labels y_test_encoded, then print the accuracy.