TY - JOUR
T1 - Prototype optimization for nearest neighbor classifiers using a two-layer perceptron
AU - Yan, Hong
PY - 1993/2
Y1 - 1993/2
N2 - The performance of a nearest neighbor classifier is degraded if only a small number of training samples are used as prototypes. An algorithm is presented for modifying the prototypes so that the classification rate can be increased. This algorithm makes use of a two-layer perceptron with one second-order input. Each hidden node of the perceptron represents a prototype and the weights of connections between a hidden node and the input nodes are initially set equal to the feature values of the corresponding prototype. The weights are then changed using a gradient-based algorithm to generate a new prototype. The algorithm has been tested with good results. © 1993.
AB - The performance of a nearest neighbor classifier is degraded if only a small number of training samples are used as prototypes. An algorithm is presented for modifying the prototypes so that the classification rate can be increased. This algorithm makes use of a two-layer perceptron with one second-order input. Each hidden node of the perceptron represents a prototype and the weights of connections between a hidden node and the input nodes are initially set equal to the feature values of the corresponding prototype. The weights are then changed using a gradient-based algorithm to generate a new prototype. The algorithm has been tested with good results. © 1993.
KW - Multi-layer neural networks
KW - Nearest neighbor classifiers
KW - Prototypes
KW - Recognition rate
KW - Training of a perceptron
UR - http://www.scopus.com/inward/record.url?scp=0027539938&partnerID=8YFLogxK
UR - https://www.scopus.com/record/pubmetrics.uri?eid=2-s2.0-0027539938&origin=recordpage
U2 - 10.1016/0031-3203(93)90040-4
DO - 10.1016/0031-3203(93)90040-4
M3 - RGC 21 - Publication in refereed journal
SN - 0031-3203
VL - 26
SP - 317
EP - 324
JO - Pattern Recognition
JF - Pattern Recognition
IS - 2
ER -