Semi-Supervised Mixture-of-Experts Classification

Authors:
Grigoris Karakoulas;Ruslan Salakhutdinov
Affiliations:
University of Toronto, Canada;University of Toronto, Canada
Venue:
ICDM '04 Proceedings of the Fourth IEEE International Conference on Data Mining
Year:
2004

Citing 0
Cited 3

Learning from labeled and unlabeled data: an empirical study across techniques and domains

Journal of Artificial Intelligence Research
Global/local hybrid learning of mixture-of-experts from labeled and unlabeled data

HAIS'11 Proceedings of the 6th international conference on Hybrid artificial intelligent systems - Volume Part I
Improved generative semisupervised learning based on finely grained component-conditional class labeling

Neural Computation

Quantified Score

Hi-index	0.00

Visualization

Abstract

We introduce a mixture-of-experts technique that is a generalization of mixture modeling techniques previously suggested for semi-supervised learning. We apply the bias-variance decomposition to semi-supervised classification and use the decomposition to study the effects from adding unlabeled data when learning a mixture model. Our empirical results indicate that the biggest gain from adding unlabeled data comes from the reduction of the model variance, whereas the behavior of the bias error term heavily depends on the correctness of the underlying model assumptions.