Sciweavers

Free Online Productivity Tools i2Speak i2Symbol i2OCR iTex2Img iWeb2Print iWeb2Shot i2Type iPdf2Split iPdf2Merge i2Bopomofo i2Arabic i2Style i2Image i2PDF iLatex2Rtf Sci2ools

10

IJON
2006

favoriteEmaildiscussreport

99views more IJON 2006»

Learning vector quantization: The dynamics of winner-takes-all algorithms

13 years 4 months ago

Learning vector quantization: The dynamics of winner-takes-all algorithms

Download www.cs.rug.nl

Winner-Takes-All (WTA) prescriptions for Learning Vector Quantization (LVQ) are studied in the framework of a model situation: Two competing prototype vectors are updated according to a sequence of example data drawn from a mixture of Gaussians. The theory of on-line learning allows for an exact mathematical description of the training dynamics, even if an underlying cost function cannot be identified. We compare the typical behavior of several WTA schemes including basic LVQ and unsupervised Vector Quantization. The focus is on the learning curves, i.e. the achievable generalization ability as a function of the number of training examples.

Michael Biehl, Anarta Ghosh, Barbara Hammer

Real-time Traffic

Exact Mathematical Description | IJON 2006 | Unsupervised Vector Quantization | Vector Quantization |

claim paper

Related Content

» Learning dynamics and robustness of vector quantization and neural gas

» Phase transitions in Vector Quantization

» Rapid Online Learning of Objects in a Biologically Motivated Recognition Architecture

» Extensions of vector quantization for incremental clustering

» Dynamic Hyperparameter Scaling Method for LVQ Algorithms

» Learning Vector Quantization generalization ability and dynamics of competing prototypes

» Alternative learning vector quantization

» A Note on Learning Vector Quantization

» A New Initial Codebook Algorithm for Learning Vector Quantization

Post Info
More Details (n/a)

Added	12 Dec 2010
Updated	12 Dec 2010
Type	Journal
Year	2006
Where	IJON
Authors	Michael Biehl, Anarta Ghosh, Barbara Hammer

Comments (0)