We just concluded the Scikit-Learn-based unsupervised learning portion of the book. Now we will move to neural network-based unsupervised learning. In the next few chapters, we will introduce neural networks, including the popular frameworks used to apply them, TensorFlow and Keras.
In Chapter 7, we will use an autoencoder—a shallow neural network—to automatically perform feature engineering and feature selection. Moving on from there, in Chapter 8, we will apply autoencoders to a real-world problem. Following that, Chapter 9 explores how to turn unsupervised learning problems into semisupervised ones, leveraging the few labels we have to improve the precision and recall of a purely unsupervised model.
Once we are finished reviewing shallow neural networks, we will look at deep neural networks in the last portion of the book.