In the context of signal classification, this paper assembles and compares criteria to easily judge the discrimination quality of a set of feature vectors. The quality measures are based on the assumption that a Support Vector Machine is used for the final classification. Thus, the ultimate criterion is a large margin separating the two classes. We apply the criteria to control the feature extr...

We consider Support Vector Machines (SVMs) learned together with linear transformations of the feature spaces on which they are applied. Under this scenario the radius of the smallest data enclosing sphere is no longer fixed. Therefore optimizing the SVM error bound by considering both the radius and the margin has the potential to deliver a tighter error bound. In this paper we present two nov...

We consider a simple model for the strangeness radius of the nucleon. The model is based on vector meson dominance (VMD) and ω − φ mixing in addition to a kaon cloud contribution. We find that the VMD contribution is similar in magnitude and of the same sign as the kaon contribution to the Sachs strangeness radius and is significantly larger than the kaon contribution to the Dirac radius.

The concept of radius of generalization was introduced in CC3 and thus this neural network overcame the generalization problem that plagued the earlier CC2 network. The Hamming distance was used for classification between binary vectors, i.e. any test vector whose Hamming distance from a training vector is smaller than the radius of generalization of the network is classified in the same output...

In traditional fuzzy support vector machine(FSVM), membership function is established in global scope will reduce the membership of support vectors, and the FSVM based dismissing margin increases the training speed, but will remove some support vector artificially. So, a new algorithm of Fuzzy Support Vector Machine with Dual Membership based on Hypersphere (HDM-FSVM) is proposed. In this algor...

In this note a simple formula for the real stability radius of uncertain positive linear continuous time systems is established and it is shown that the real stability radius coincides with the complex one. Arbitrary disturbance norms induced by monotonic vector norms (e.g. p-norms, 1 p 1) are considered. The distance of intervals of positive systems from instability is also determined.

A new algorithm for Support Vector regression is described. For a priori chosen 1/, it automatically adjusts a flexible tube of minimal radius to the data such that at most a fraction 1/ of the data points lie outside. Moreover, it is shown how to use parametric tube shapes with non-constant radius. The algorithm is analysed theoretically and experimentally.

Soluble model of a relativistic particle describing a bag of matter with fixed radius held together in perfect balance by a self-consistent combination of three forces generated by electromagnetic and massive scalar and vector fields is presented. For realistic values of parameters the bag radius becomes that of a proton.

The Newton radius of a code is the largest weight of a uniquely correctable error. The covering radius is the largest distance between a vector and the closest codeword. A couple of relations involving the Newton and covering radii are discussed. c © 2001 Elsevier Science B.V. All rights reserved.