A New Categorization Numerical Scheme for Mobile Robotic Computing Using Odor Data-set Recognition as a Case

Categorization is one of the most active research and application areas of Data Mining. In this paper, we address the problem of pattern categorization in mobile robotic computing. It is the task of automatically sorting a set of patterns into categories from a predefined set. Most categorization algorithms are sensitive to noise, architecture configuration, Bellman's curse of dimensionality, instability, and complex shapes. Hence, in the present study, a novel numerical scheme (RC) for pattern categorization which provides a good generalization ability with a small empirical error, is described. The experimental study with E-nose of six different MOX gas sensors is presented. Our evaluation method demonstrates the effectiveness and multidisciplinary applications of our approach.