A modified active subset selection method based on quadratic Rényi entropy and a fast cross-validation for fixed-size least-squares support vector machines is proposed for classification and regression with optimized tuning process. The kernel bandwidth of the entropy based selection criterion is optimally determined according to the solve-the-equation plug-in method. Also a fast cross-validati...

In this paper, we propose a fast and accurate approximation to the information potential of Information Theoretic Learning (ITL) using the Fast Gauss Transform (FGT). We exemplify here the case of the Minimum Error Entropy criterion to train adaptive systems. The FGT reduces the complexity of the estimation from O(N) to O(pkN) where p is the order of the Hermite approximation and k the number o...

This paper studies convergence to equilibrium for the spatially inhomogeneous linear relaxation Boltzmann equation in entropy and related functionals, $p$-entropies. Villani [Hypocoercivity, Memoirs of American Mathematical Society, Vol. 202, Providence, RI, 2006] proved entropic hypocoercivity a class PDEs Hörmander sum-of-squares form. It was an open question prove such result operator which ...

A single-parameter harmonic Hamiltonian based on local packing density and contact topology is proposed for studying residue fluctuations in native proteins. The internal energy obeys an equipartition law, and free energy changes result from entropy fluctuations only. Frequency–wave-number maps show communication between residues involved in slow and fast modes. Fast modes are strongly localize...

This paper presents a novel approach to automatic classifying and identifying of tree leaves using image segmentation fusion. With the development of mobile devices and remote access, automatic plant identification in images taken in natural scenes has received much attention. Image segmentation plays a key role in most plant identification methods, especially in complex background images. Wher...

the estimate of entropy (sample entropy), has been introduced by vasicek (1976), for the first time. in this paper, we provide an estimate of entropy of order statistics, that is the extention of the entropy estimate. then we present an application of the entropy estimate of order statistics as a test statistic for symmetry of distribution versus skewness. the proposed test has been compared wi...

We show that any encryption/decryption system can be converted to an algorithm which uses roughly three machine operations per bit. We prove that our technique yields an encryption/decryption system which is as secure as the original system provided the plaintext has full-entropy. We then show how to remove the full-entropy assumption. Our techniques speed up software implementations of RSA by ...

Non-negative matrix based language models have been recently introduced [1] as a computationally efficient alternative to other feature-based models such as maximum-entropy models. We present a new entropy based pruning algorithm for this class of language models, which is fast and scalable. We present perplexity and word error rate results and compare these against regular n-gram pruning. We a...

Recursive interleaved entropy coding is a promising new adaptable binary entropy coding technique that offers fast encoding and decoding at high compression efficiency [1]. However, with the encoding methods presented in [1], the maximum memory required by the encoder is proportional to the source sequence length. In this article, we present a new encoding technique, and corresponding decoding ...