Chapter 10: Deploying a Deep Learning Network
In the previous sections of this book, we covered the training of deep neural networks for many different use cases, starting with an autoencoder for fraud detection, through Long Short-Term Memory (LSTM) networks for energy consumption prediction and free text generation, all the way to cancer cell classification. But training the network is not the only part of a project. Once a deep learning network is trained, the next step is to deploy it.
During the exploration of some of the use cases, a second workflow has already been introduced, to deploy the network to work on real-world data. So, you have already seen some deployment examples. In this last section of the book, however, we focus on the many deployment options for machine learning models in general, and for trained deep learning networks in particular.
Usually, a second workflow is built and dedicated to deployment. This workflow reads the trained model and the new real...
