7. Convolutional Neural Networks

Activity 7.01: Building a CNN with More ANN Layers

Solution:

There are several possible ways to arrive at a solution for this activity. The following steps describe one of these methods and are similar to those used on the CIFAR-10 dataset earlier in the chapter:

1. Start a new Jupyter notebook.
2. Import the TensorFlow library:

   import tensorflow as tf

3. Import the additional libraries needed:

   import numpy as np
   import matplotlib.pyplot as plt
   import tensorflow as tf
   import tensorflow_datasets as tfds
   from tensorflow.keras.layers import Input, Conv2D, Dense, Flatten, \
       Dropout, Activation, Rescaling
   from tensorflow.keras.models import Model
   from sklearn.metrics import confusion_matrix, ConfusionMatrixDisplay

4. Load the CIFAR-100 dataset directly from tensorflow_datasets and view its properties:

   (c100_train_dataset, c100_test_dataset), \
   dataset_info = tfds.load('cifar100',\
        ...