Topology for Time Series
Topology studies the global properties of a space, such as the data collected from a system changing over time, and can classify objects by those properties. This makes it very useful for comparing a group of systems that may or may not have the same properties. By learning about how topology can be used to study time series, you’ll be equipped to study data that changes over time at a global system level.
This talk will introduce participants to topological algorithms to compare time series data. We’ll overview time series data, examine South African and Egyptian stock market trends, learn about topology and an algorithm called persistent homology, and implement our analysis in R. Participants will come away with an understanding of the persistent homology algorithm, an understanding of the caveats of analyzing/comparing time series data, and an understanding of how to implement the algorithm in R.