@inproceedings {pub2489,
	title = {Rejection strategies for Learning Vector Quantization - a comparison of probabilistic and deterministic approaches},
	author = {Heiko Wersing AND David Nebel AND Barbara Hammer AND Thomas Villmann AND Lydia Fischer},
	year = {2014},
	month = {April},
	abstract = {Reject options in multiclass classification schemes enable a direct man-machine interaction which allow an improvement of the classification task with human expert knowledge in regions with class overlap or outliers.
We compare two rejection strategies which are derived from Generalized Learning Vector Quantization with statistical motivated strategies.
These two options are heuristics only but with less additional computational effort on these approaches.
The statistical counterpart for Generalized Learning Vector Quantization approaches is the Robust Soft Learning Vector Quantization approach which provides a trained Gaussian mixture model with probabilities.
These probabilities can be interpreted as confidences and can be applied as reject option directly.
The original Robust Soft Learning Vector Quantization approach uses a discriminative Gaussian mixture model which does not necessarily model the data density. 
Another version of Robust Soft Learning Vector Quantization has a generative component as well and is therefore more suited in providing probabilities as confidence values. In this paper we compare these two heuristic reject options against the statistical founded approaches of the two versions of Robust Soft Learning Vector Quantization.},
	publisher = {Springer},
	booktitle = {Advances in Self-Organizing Maps and Learning Vector Quantization - Proceedings of the 10th International Workshop, WSOM},
	editor = {Thomas Villmann and Frank-Michael Schleif and Marika Kaden and Mandy Lange},
	city = {Mittweida, Germany},
	volume = {295},
	pages = {109-118},
	series = {Advances in Intelligent Systems and Computing}
}