Test feature classifiers: performance and application

In this paper, we present class of combinatorical-logical classifiers called test feature classifiers. Introducing kernels and a rejection option we discuss the properties and performance of the proposed classifiers. To test the performance of the classifiers, we apply them to both artificial and real data, and compare them with conventional ones. Experimental results show that the proposed classifiers yield a higher recognition rate than conventional ones, have a high ability of generalization, and suggest that they can be used in a variety of pattern recognition applications