Wlodzislaw Duch 
Neural minimal distance methods, 
Third Conference on Neural Networks and Their Applications, 
Kule, Poland, October 1997, pp. 183-188
	
Minimal distance methods are simple and in some circumstances highly accurate. In
this paper relations between neural and minimal distance methods are
investigated. Neural realization facilitates new versions of minimal distance
methods. In k-NN (k-nearest-neighbor) method only one parameter is optimized. In
NN-r approach k is variable but the radius r is optimized.
Parametrization of distance functions, distance-based weighting of neighbors,
active selection of reference vectors from the training set and relations to the
case-based reasoning are also discussed.

