An adaptive algorithm for training of a nearest neighbour (NN) classifier is developed in this paper. This learning rule has some similarity to the well-known LVQ method, but uses the nearest centroid neighbourhood concept to estimate optimal locations of the codebook vectors. The aim of this approach is to improve the performance of the standard LVQ algorithms when using a very small codebook. The behaviour of the learning technique proposed here is experimentally compared to those of the plain k-NN decision rule and the LVQ algorithms.
Learning vector quantization with alternative distance criteria
1999-01-01
75770 byte
Conference paper
Electronic Resource
English
Learning Vector Quantization with Alternative Distance Criteria
British Library Conference Proceedings | 1999
|Vector Quantization Based on Dynamic Adjustment of Mahalanobis Distance
British Library Conference Proceedings | 1996
|Vector quantization with distance constraints for enhanced post-processing [2669-03]
British Library Conference Proceedings | 1996
|