Skip to collection list
Skip to video grid
Topological Data Analysis with Persistent Homology
Bryn Keller, Senior Data Scientist, Intel
Persistent homology is one of several tools from algebraic topology that's been applied to data analysis. It suggests a strategy for dealing with difficult problems with shapes, such as the shapes of decision boundaries and data manifolds. We begin with an accessible review of persistent homology and then draw the connections between deep learning and persistent homology that may help to strengthen both, using examples from neuroscience and medicinal chemistry.