Topological data analysis from a practical and mathematical perspective
- 日時
- 2020年7月15日(水)16:00 - 18:10 (JST)
- 講演者
-
- 池 祐一 (株式会社富士通研究所 人工知能研究所 研究員)
- 会場
- via Zoom
- 言語
- 英語
1. Topological data analysis and its applications
In this talk, I will explain some methods in topological data analysis (TDA) and their applications. First I recall persistent homology, which is a central tool to analyze the "shape" of a point cloud set. Then I show several applications to material science and time-series analysis. I also talk about our collaborative research with Inria on noise-robust persistent homology and an automated vectorization method of persistence diagrams.
2. Persistence-like distance on sheaf category and displacement energy
In this talk, I will talk about relation among sheaf theory, persistence modules, and symplectic geometry. We introduce a persistence-like distance on Tamarkin sheaf category and prove a stability result with respect to Hamiltonian deformation of sheaves. Based on this result, we propose a new sheaf-theoretic method to give a lower bound of the displacement energy of compact subsets of a cotangent bundle.
This is a joint work with Tomohiro Asano.
*Please contact Keita Mikami's mail address to get access to the Zoom meeting room.