We study parallel two-level overlapping Schwarz algorithms for solving nonlinear nite element problems, in particular, for the full potential equation of aerodynamics discretized in two dimensions with bilinear elements. The overall algorithm, Newton-Krylov-Schwarz (NKS), employs an inexact nite-di erence Newton method and a Krylov space iterative method, with a two-level overlapping Schwarz me...

In this paper we describe the use of Rademacher penalization for model selection. As in Vapnik's Guaranteed Risk Minimization (GRM), Rademacher penalization attemps to balance the complexity of the model with its t to the data by minimizing the sum of the training error and a penalty term, which is an upper bound on the absolute di erence between the training error and the generalization error....

Abstract. Topological tools, such as Conley index theory, have inspired rigorous computational methods for studying dynamics. These methods rely on the construction of an outer approximation, a combinatorial representation of the system that incorporates small, bounded error. In this work, we present an automated approach for constructing outer approximations for systems in a class of integrodi...

A computational model of auditory localization based on processing spectral amplitude cues for localizing broadband high frequency sound sources is presented. The cues extracted are binaural spectral level di erence patterns of the Head Related Transfer Functions (HRTF) corresponding to the direction of the sound source. Four di erent pattern classi ers are used to evaluate the feasibility of l...

We present the detailed analysis of the spontaneous emission coupling factor ( factor) of the microcavity based on a 2D photonic crystal in an optically thin dielectric slab. We investigate the maximum value that can be achieved with this microcavity and discuss its dependence on the quantum well position, as well as on the pumping area diameter. The analysis is performed using the general meth...

This paper evaluates the Collective Neuro-Evolution (CONE) method, comparative to a related controller design method, in a simulated multi-robot system. CONE solves collective behavior tasks, and increases task performance via facilitating behavioral specialization. Emergent specialization is guided by genotype and behavioral specialization di erence metrics that regulate genotype recombination...

Computation of shortest paths on free-form surfaces is an important problem in ship design, robot motion planning, computation of medial axis transforms of trimmed surface patches, terrain navigation and NC machining. The objective of this paper is to provide an e cient and reliable method for computing the shortest path between two points on a free-form parametric surface and the shortest path...

It has been widely observed that Krylov space solvers based on two three term re currences can give signi cantly less accurate residuals than mathematically equivalent solvers implemented with three two term recurrences In this paper we attempt to justify this di erence theoretically by analyzing the gap between the recursively and the explicitly computed residuals It is shown that in contrast ...

| 3-D shape modeling is very important for e cient shape description and recognition. Superquadrics that is a parametric 3-D shape modeling function can represent various shapes by using a single equation with some parameters. In this study, the superquadrics parameters of 3-D shape are estimated from a 2-D shading image by using genetic algorithm (GA), which is an optimizing technique based on...

This paper considers the security of iterated block ciphers against the di erential cryptanalysis introduced by Biham and Shamir. Di erential cryptanalysis is a chosen-plaintext attack on secret-key block ciphers that are based on iterating a cryptographically weak function r times (e.g., the 16-round Data Encryption Standard (DES) ). It is shown that the success of such attacks on an r-round c...