Distributed support vector machines over dynamic balanced directed networks

In this letter, we consider the binary classification problem via distributed Support Vector Machines (SVMs), where the idea is to train a network of agents, with limited share of data, to cooperatively learn the SVM classifier for the global database. Agents only share processed information regarding the classifier parameters and the gradient of the local loss functions instead of their raw data. In contrast to the existing work, we propose a continuous-time algorithm that incorporates network topology changes in discrete jumps. This hybrid nature allows us to remove chattering that arises because of the discretization of the underlying CT process. We show that the proposed algorithm converges to the SVM classifier over time-varying weight balanced directed graphs by using arguments from the matrix perturbation theory.

Description

Publisher Copyright: IEEE

Keywords

Distributed databases, distributed optimization, Heuristic algorithms, Manganese, matrix perturbation theory., Radio frequency, Signal processing algorithms, Support vector machines, Support Vector Machines, Switches

DOI

10.1109/LCSYS.2021.3086388

Citation

Doostmohammadian , M , Aghasi , A , Charalambous , T & Khan , U A 2022 , ' Distributed support vector machines over dynamic balanced directed networks ' , IEEE Control Systems Letters , vol. 6 , 9446550 , pp. 758-763 . https://doi.org/10.1109/LCSYS.2021.3086388

Permanent link to this item

https://urn.fi/URN:NBN:fi:aalto-202106167413

Collections

[article-cris] Sähkötekniikan korkeakoulu / ELEC

Show all metadata