On distributed online classification in the midst of concept drifts

Authors:
Zaid J. Towfic;Jianshu Chen;Ali H. Sayed
Affiliations:
Electrical Engineering Department, University of California, Los Angeles, CA 90095, USA;Electrical Engineering Department, University of California, Los Angeles, CA 90095, USA;Electrical Engineering Department, University of California, Los Angeles, CA 90095, USA
Venue:
Neurocomputing
Year:
2013

Citing 13
Cited 0

The nature of statistical learning theory

The nature of statistical learning theory
Learning in the presence of concept drift and hidden contexts

Machine Learning
Adaptive filter theory (3rd ed.)

Adaptive filter theory (3rd ed.)
The anatomy of a large-scale hypertextual Web search engine

WWW7 Proceedings of the seventh international conference on World Wide Web 7
Convex Optimization

Convex Optimization
Matrix Analysis For Scientists And Engineers

Matrix Analysis For Scientists And Engineers
Adaptive Filters

Adaptive Filters
Dynamic Weighted Majority: An Ensemble Method for Drifting Concepts

The Journal of Machine Learning Research
LIBLINEAR: A Library for Large Linear Classification

The Journal of Machine Learning Research
Pattern Recognition, Fourth Edition

Pattern Recognition, Fourth Edition
Diffusion LMS strategies for distributed estimation

IEEE Transactions on Signal Processing
On classifying drifting concepts in P2P networks

ECML PKDD'10 Proceedings of the 2010 European conference on Machine learning and knowledge discovery in databases: Part I
Diffusion Least-Mean Squares Over Adaptive Networks: Formulation and Performance Analysis

IEEE Transactions on Signal Processing - Part II

Quantified Score

Hi-index	0.01

Visualization

Abstract

In this work, we analyze the generalization ability of distributed online learning algorithms under stationary and non-stationary environments. We derive bounds for the excess-risk attained by each node in a connected network of learners and study the performance advantage that diffusion has over individual non-cooperative processing. We conduct extensive simulations to illustrate the results.