Summary

In this chapter, we understood how to implement a convolution neural network classifier in Keras. You now have a brief understanding of what convolution, average, max pooling, and dropout are, and you also built a deep model. You understood how to reduce overfitting as well as how to generate more/validation in data to build a generalizable model when you have less data than you need. Finally, we assessed the model's performance on test data and determined that we succeeded in achieving our goal. We ended this chapter by introducing you to autoencoders.