Ensemble learning in linearly combined classifiers via negative correlation

Authors:
Manuela Zanda;Gavin Brown;Giorgio Fumera;Fabio Roli
Affiliations:
School of Computer Science, University of Manchester, UK;School of Computer Science, University of Manchester, UK;Dept of Electrical and Electronic Engineering, University of Cagliari, Italy;Dept of Electrical and Electronic Engineering, University of Cagliari, Italy
Venue:
MCS'07 Proceedings of the 7th international conference on Multiple classifier systems
Year:
2007

Citing 4
Cited 2

Neural networks and the bias/variance dilemma

Neural Computation
Ensemble learning via negative correlation

Neural Networks
A Theoretical and Experimental Analysis of Linear Combiners for Multiple Classifier Systems

IEEE Transactions on Pattern Analysis and Machine Intelligence
Managing Diversity in Regression Ensembles

The Journal of Machine Learning Research

An Anticorrelation Kernel for Subsystem Training in Multiple Classifier Systems

The Journal of Machine Learning Research
Maps ensemble for semi-supervised learning of large high dimensional datasets

ISMIS'08 Proceedings of the 17th international conference on Foundations of intelligent systems

Quantified Score

Hi-index	0.00

Visualization

Abstract

We investigate the theoretical links between a regression ensemble and a linearly combined classification ensemble. First, we reformulate the Tumer & Ghosh model for linear combiners in a regression context; we then exploit this new formulation to generalise the concept of the "Ambiguity decomposition", previously defined only for regression tasks, to classification problems. Finally, we propose a new algorithm, based on the Negative Correlation Learning framework, which applies to ensembles of linearly combined classifiers.