Estimation of the number of signals from features of the covariancematrix: a supervised approach

The purpose of this paper is to provide a fast and simplified detection test for use in the presence of a small number of sources (from 0-2), which is able to accommodate correlated paths and nonwhite noise; conventional eigenvalue-based criteria are unable to do so. For a uniform linear array, using common sense arguments, a small set of significant features of the covariance matrix are used as inputs to a neural net. The nonlinear transfer function of the neural net is adjusted by supervised training to provide the discriminant functions for order selection in its outputs. Results from the net are then compared with conventional criteria and demonstrate superior performance, in particular, for correlated sources and small sample sizes. Training may be introduced for known nonwhite noise, which serves to maintain high performance for reasonable correlation lengths     

Preservation of protective capacity of hyperimmune anti-Stx2 bovine colostrum against enterohemorrhagic Escherichia coli O157:H7 pathogenicity after pasteurization and spray-drying processes

Enterohemorrhagic Escherichia coli (EHEC) O157:H7 is a major etiologic agent that causes bloody diarrhea, hemorrhagic colitis, and hemolytic uremic syndrome (HUS). Shiga toxin (Stx) is the main virulence factor of EHEC responsible for the progression to HUS. Although many laboratories have made efforts to develop an effective treatment for Stx-mediated HUS, a specific therapy has not been found yet. Human consumption of bovine colostrum is known to have therapeutic effects against several gastrointestinal infections because of the peptide and proteins (including antibodies) with direct antimicrobial and endotoxin-neutralizing effects contained in this fluid. We have previously demonstrated that colostrum from Stx type 2 (Stx2)-immunized pregnant cows effectively prevents Stx2 cytotoxicity and EHEC O157:H7 pathogenicity. In this study we evaluated the preservation of the protective properties of hyperimmune colostrum against Stx2 (HIC-Stx2) after pasteurization and spray-drying processes by performing in vitro and in vivo assays. Our results showed that reconstituted HIC-Stx2 colostrum after pasteurization at 60°C for 60 min and spray-dried under optimized conditions preserved specific IgG that successfully neutralized Stx2 cytotoxicity on Vero cells. Furthermore, this pasteurized/dehydrated and reconstituted HIC-Stx2 preserved the protective capacity against EHEC infection in a weaned mice model. The consumption of hyperimmune HIC-Stx2 bovine colostrum could be effective for HUS prevention in humans as well as in EHEC control in calves. However, further studies need to be done to consider its use for controlling EHEC infections.     

The Empirical Likelihood method applied to covariance matrix estimation

This paper presents a new estimation scheme for signal processing problems in unknown noise field. The Empirical Likelihood has been introduced in the mathematical community, but, surprisingly, it is still unknown in the signal processing community. This estimation method is an alternative to estimate unknown parameters without using a model for the probability density function.The aim of this paper is twofold: first, the Empirical Likelihood theory is presented and revisited thanks to the moment method. Its properties are derived. Second, to emphasize all the potentiality of this method, we address the problem of Toeplitz matrix estimation: this leads us to obtain improved estimates in comparison to conventional ones, as shown in simulations.     

Initialization of Supervised Training for Parametric Estimation

This paper concerns the initialization problem of the training algorithm in Neural Networks. We focus herein on backpropagation networks with one hidden layer. The initialization of the weights is crucial; if the network is incorrectly initialized, it converges to local minima. The classical random initialization therefore appears as a very poor solution. If we were to consider the Taylor development of the mapping problem and the nonlinearity of sigmoids, the improvements could be very significant. We propose a new initialization scheme based on the search for an explicit approximate solution to the problem of mapping between pattern and target. Simulation results are presented which show that these original initializations avoid local minima, reduce training time, obtain a better generalization and estimate the network's size.     

Statistical prediction of monopulse angle measurement

The statistical prediction of monopulse angle measurement where the amplitude fluctuation law of the signal source is unknown and multiple observations are available is presented. The mean and the variance of the monopulse ratio have been derived for the general case where the signal source is embedded in spatially coloured Gaussian noise and the receiver sum-channel is possibly completed by a detection step. This result completes the statistical prediction of a Swerling target and provides a powerful method to compute the mean and variance of the monopulse ratio conditioned by the detection test, whatever the amplitude fluctuation law of the signal source.     

On the influence of a detection step on lower bounds for deterministic parameter estimation

A wide variety of actual processing requires a detection step, whose main effect is to restrict the set of observations available for parameter estimation. Therefore, as a contribution to the theoretical formulation of the joint detection and estimation problem, we address the derivation of lower bounds for deterministic parameters conditioned by a binary hypothesis testing problem. The main result is the introduction of a general scheme-detailed in the particular case of CRB-enabling the derivation of conditional deterministic MSE lower bounds. To prove that it is meaningful, we also show, with the help of a fundamental application, that the problem of lower bound tightness at low SNR may arise from an incorrect lower bound formulation that does not take into account the true nature of the problem under investigation: a joint detection-estimation problem.