Research work · 2014
About model-based distances
Mattioli Gabriel
Frontiers in Artificial Intelligence and Applications
Abstract
The literature of any scientifical field is full of models. In this work it will be shown how every model defines a distance function in a natural way. This distance is much more suitable for specific problems than the ones usually used (generally Euclidean or Manhattan), because its directly related with the particular problem and does not has to be chosen arbitrarily.