Research Seminars: Geometry & Topology

Fall 2026

Time & Location: All talks are on Monday in Gibson Hall ___ at _:00 PM unless otherwise noted.
Organizer: Komendarczyk, Rafal

Archives

 

 

August 27, 2026 

Geometry & Topology

Title: Persistent Homology Learning Seminar

Speaker: Rafal Komendarczyk - Tulane

Abstract: We would like to gauge interest and work out logistics before committing to a regular schedule.

Please join us to discuss:

     - Current status and topics covered so far
     - Interest in continuing as a biweekly seminar
     - Scheduling preferences

**Course Registration:** This seminar is offered as MATH-7760-02 (Persistent Homology) for 3 graduate credits it will meet on Thursdays 12:30 PM – 1:30 PM in Norman Mayer MA-104. The course just appeared in the roster and registration remains open through Friday, August 28, 2026. Students interested in registering for credit should do so before the deadline.
 

Location: Norman Mayer MA-104

Time: 12:30 PM

_______________

September 03, 2026 

Geometry & Topology

Title: Persistent Homology Learning Seminar

Speaker: Rafal Komendarczyk - Tulane University

Abstract: The Persistent Homology Learning Seminar (MATH-7760-02) resumes this week and meets weekly through the fall, Thursdays 12:30-1:30 PM in Gibson Hall GI-308.
Please note the room: this is not where we met on August 27.

The seminar is organized around one question: when are two metric spaces close, and which topological invariants respect that closeness? We take the
Gromov-Hausdorff distance as the measure of similarity between metric spaces and work toward the stability theorem

d_B(B(X), B(Y)) <= 2 d_GH(X, Y), which says that the persistence barcode is a Lipschitz invariant of the data.

All interested are welcome to attend. No prior exposure to persistent homology is assumed.
 

Location: Gibson Hall GI-308

Time: 12:30 PM

_______________

 
September 10, 2026 

Geometry & Topology

Title: Persistent Homology Learning Seminar (LSC)

Speaker: Rafal Komendarczyk - Tulane University

Abstract: Last week d_GH was defined as an infimum of Hausdorff distances over all isometric embeddings of two compact metric spaces into a common ambient space: conceptually clean, but an infimum over a proper class. This week we prove the correspondence formula d_GH(X,Y) = (1/2) inf_C dis(C), the infimum taken over correspondences C in X x Y. It removes the ambient space entirely and is the form in which d_GH enters every later estimate.
 

---------------------------


The seminar is organized around one question: when are two metric spaces close, and which topological invariants respect that closeness? We take the Gromov-Hausdorff distance as the measure of similarity between metric spaces and work toward the stability theorem d_B(B(X), B(Y)) <= 2 d_GH(X, Y), which says that the persistence barcode is a Lipschitz invariant of the data. 

Location: Gibson Hall GI-308

Time: 12:30 PM

_______________