Abstract
Exploring multidimensional data is a complex analytic task. We propose a generative model called Generative Simplicial Complex, to extract topological invariants called Betti numbers from the data. The GSC is used to analyze toys data and image data. The GSC appears to be more robust to noise than the Witness Complex, a state of the art geometrical technique to extract Betti numbers of point set data.
Translated title of the contribution | Extraction of Betti numbers with a generative model |
---|---|
Original language | French |
Pages (from-to) | 97-102 |
Number of pages | 6 |
Journal | Revue des Nouvelles Technologies de l'Information |
Volume | E.24 |
Publication status | Published - 2013 |
Externally published | Yes |
Event | 13emes Journees Francophones sur l'Extraction et la Gestion des Connaissances, EGC 2013 - 13th French-Speaking Conference on Knowledge Discovery and Management, EGC 2013 - Toulouse, France Duration: 29 Jan 2013 → 1 Feb 2013 |