Summary

In this chapter, we covered the process of dimensionality reduction and PCA. We completed a number of exercises and developed the skills to reduce the size of a dataset by extracting only the most important components of variance within the data, using both a manual PCA process and the model provided by scikit-learn. During this chapter, we also returned the reduced datasets back to the original dataspace and observed the effect of removing the variance on the original data. Finally, we discussed a number of potential applications for PCA and other dimensionality reduction processes. In our next chapter, we will introduce neural network-based autoencoders and use the Keras package to implement them.