9 Autoencoders
This chapter covers
- Introducing autoencoders
- Training of autoencoders
- Types of autoencoders
- Python code using TensorFlow and Keras
Out of intense complexities, intense simplicities emerge.
In the preceding chapter, we explored the concepts of deep learning. In this chapter, we start with unsupervised deep learning. Autoencoders are the very first topic. We will first cover the basics of autoencoders, what are they, and how we train them. We then get into the different types of autoencoders followed by a Python code on the implementation. Welcome to the ninth chapter, and all the very best!
9.1 Technical toolkit
We will continue to use the same version of Python and Jupyter Notebook as we have used so far. The codes and datasets used in this chapter have been checked in at the GitHub location. You need to install a couple of Python libraries in this chapter: tensorflow and keras.
9.2 Feature learning
Predictive modeling is quite an interesting topic. Across various domains and business functions, predictive modeling is used for various purposes like predicting the sales for a business in the next year, the amount of rainfall expected, whether the incoming credit card transaction is fraud or not, whether the customer will make a purchase or not, and so on. The use cases are many, and all the aforementioned use cases fall under supervised learning algorithms.