Titlebar

Export bibliographic data
Literature by the same author
plus on the publication server
plus at Google Scholar

 

Quantitative Robustness of Localized Support Vector Machines

Title data

Dumpert, Florian:
Quantitative Robustness of Localized Support Vector Machines.
Bayreuth , 2019

Official URL: Volltext

Abstract in another language

The huge amount of available data nowadays is a challenge for kernel-based machine learning algorithms like SVMs with respect to runtime and storage capacities. Local approaches might help to relieve these issues and to improve statistical accuracy. It has already been shown that these local approaches are consistent and robust in a basic sense. This article refines the analysis of robustness properties towards the so-called influence function which expresses the differentiability of the learning method: We show that thereis a differentiable dependency of our locally learned predictor on the underlying distribution. The assumptions of the proventheorems can be verified without knowing anything about this distribution. This makes the results interesting also from an applied point of view.

Further data

Item Type: Preprint, postprint
Keywords: Machine learning; localized learning; robustness; influence function
Institutions of the University: Faculties > Faculty of Mathematics, Physics und Computer Science > Department of Mathematics > Chair Mathematics VII - Stochastics > Chair Mathematics VII - Stochastics - Univ.-Prof. Dr. Andreas Christmann
Result of work at the UBT: Yes
DDC Subjects: 500 Science > 510 Mathematics
Date Deposited: 06 Jun 2019 06:28
Last Modified: 06 Jun 2019 06:28
URI: https://eref.uni-bayreuth.de/id/eprint/49271