Bubble Algorithm for the Reduction of Reference

A vast majority of algorithms for the condensation of the reference set requires a great number of computations in case of processing a very large set, one that contains several dozens of objects. This fact formed the grounds for the presented attempt to develop a completely new classifier, an algorithm which would not only maintain the quality of classification similar to one obtained with the primary reference set but also allow to accelerate computations considerably. The proposed solution consists in covering the primary reference set with disjoint hyperspheres; however, these hyperspheres may contain objects from one class only. Classification is completed when it is determined that the classified point belongs to one of the mentioned spheres. If an object does not belong to any hypersphere, it is counted among the objects of the same class, to which the objects from the nearest hypersphere belong (the distance to the centre of the sphere minus the radius). As was indicated by the tests, this algorithm proved to be very efficient with very large sets.