Online Distributed Learning Over Networks in RKH Spaces Using Random Fourier Features
| Autor: | Sergios Theodoridis, Pantelis Bouboulis, Symeon Chouvardas |
|---|---|
| Rok vydání: | 2018 |
| Předmět: |
FOS: Computer and information sciences
Theoretical computer science Computer science business.industry Hilbert space Contrast (statistics) 020206 networking & telecommunications Regret 02 engineering and technology Machine Learning (cs.LG) Support vector machine Computer Science - Learning Kernel (linear algebra) symbols.namesake Fourier transform Dimension (vector space) Kernel (statistics) Signal Processing 0202 electrical engineering electronic engineering information engineering symbols 020201 artificial intelligence & image processing Artificial intelligence Electrical and Electronic Engineering business Reproducing kernel Hilbert space |
| Zdroj: | IEEE Transactions on Signal Processing. 66:1920-1932 |
| ISSN: | 1941-0476 1053-587X |
| DOI: | 10.1109/tsp.2017.2781640 |
| Popis: | We present a novel diffusion scheme for online kernel-based learning over networks. So far, a major drawback of any online learning algorithm, operating in a reproducing kernel Hilbert space (RKHS), is the need for updating a growing number of parameters as time iterations evolve. Besides complexity, this leads to an increased need of communication resources in a distributed setting. In contrast, we propose to approximate the solution as a fixed-size vector (of larger dimension than the input space) using the previously introduced framework of random Fourier features. This paves the way to use standard linear combine-then-adapt techniques. To the best of our knowledge, this is the first time that a complete protocol for distributed online learning in RKHS is presented. Conditions for asymptotic convergence and boundness of the networkwise regret are also provided. The simulated tests illustrate the performance of the proposed scheme. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|