The Shape of Data

7 Tools for Topological Data Analysis

In this chapter, we’ll explore algorithms that have a direct basis in topology and use them to understand the dataset of self-reported educational data encountered in Chapter 6. The branch of machine learning that includes topology-based algorithms is called topological data analysis (TDA). You already saw some TDA in Chapter 4, where we used persistent homology to explore network differences. Persistent homology has gained a lot of attention in the machine learning community lately and has been used in psychometric data validation, image comparison analyses, pooling steps of convolutional neural networks, ...