Topological Data Analysis (TDA) is a rapidly growing field with tremendous potential for improving machine learning pipelines, unboxing the most intricate deep networks, and creating new geometric features from data. Since the knowledge barrier that protects TDA might scare some data scientists, the purpose of this track is not only to showcase TDA as a complementary tool to machine learning, but also to lower this barrier and make TDA more intuitive and accessible.
The speakers of this track are all renowned leaders in the topics they present: time series analysis using TDA, regularization of deep networks with TDA, feature engineering via TDA, and data visualisation and dimensionality reduction are the main topics discussed in this track. The subject is new, and we believe it deserves more widespread visibility within the data science community.
Concrete use-cases will be presented, in addition to more theoretical talks.