Applications of Topological Data Analysis to Data Science, Artificial Intelligence, and Machine Learning
April 28, 2022 at SDM Conference
April 28, 2022 at SDM Conference
Entropic Hyper-Connectomes Computation and Analysis by Michael G. Rawson
A Case Study on Bifurcation and Chaos with CROCKER Plots by İsmail Güzel, Elizabeth Munch, & Firas Khasawneh
DBSpan: Density-Based Clustering Using a Spanner, With an Application to Persistence Diagrams by Brittany Terese Fasy, David L. Millman, Elliott Pryor, & Nathan Stouffer
Topological Data Analysis for Anomaly Detection in Host-Based Logs by Thomas Davies
Parallel coarsening of graph data with spectral guarantees by Christopher Brissette, George Slota, & Andy Huang
TopoEmbedding, a web tool for the interactive analysis of persistent homology by Xueyi Bao, Guoxi Liu & Federico Iuricich
10:15- 10:30 AM Welcome and Morning Session Introduction
10:30 - 11:30 AM Contributed Papers 1
Topological Data Analysis for Anomaly Detection in Host-Based Logs by Thomas Davies
A Case Study on Bifurcation and Chaos with CROCKER Plots by İsmail Güzel, Elizabeth Munch, & Firas Khasawneh
TopoEmbedding, a web tool for the interactive analysis of persistent homology by Xueyi Bao, Guoxi Liu & Federico Iuricich
11:30 - 12:15 AM Panel Discussion: Advancing the Theory
Justin Curry, University at Albany SUNY
Leland McInnes, Tutte Institute for Mathematics and Computing
Bei Wang, University of Utah
12:15 - 01:30 Break
01:30 - 02:30 Keynote: Topological classification and synthesis of neuron morphologies
Kathryn Hess, Swiss Federal Institute of Technology Lausanne (EPFL)
02:30 - 03:30 Contributed Papers 2
Entropic Hyper-Connectomes Computation and Analysis by Michael G. Rawson
DBSpan: Density-Based Clustering Using a Spanner, With an Application to Persistence Diagrams by Brittany Terese Fasy, David L. Millman, Elliott Pryor, & Nathan Stouffer
Parallel coarsening of graph data with spectral guarantees by Christopher Brissette, George Slota, & Andy Huang
03:30 - 04:00 Break
04:00 - 05:00 Keynote: Topological Methods for Deep Learning
Gunnar Carlsson, Stanford University
05:00 – 06:00 Panel Discussion: Computational Tools
Erin Chambers, Saint Louis University
Greg Henselman-Petrusek, Pacific Northwest National Laboratory
Michael Lesnick, University at Albany SUNY
John Healy, Tutte Institute for Mathematics and Computing
**All times are EDT.
Panel Topics:
Advancing the theory. Open problems in the underlying theory of TDA remain and others yet to be posed. As TDA is at the intersection of pure mathematics and data science the field could potentially benefit from a conversation towards motivating advances in the theory with unsolved problems in the application domain. This panel would include theorists and practitioners.
Computational tools: There are many computational tools for TDA techniques including persistent homology, mapper, and sheaf modeling. This panel will include a discussion of state of the art as well as current gaps in our ability to apply TDA practically. Computational TDA researchers could gather ideas for open problems from the discussion of gaps while practitioners who wish to use TDA tools could gather insight from the discussion of current state of the art.