Search results for "gradient descent"
showing 10 items of 20 documents
On a global superconvergence of the gradient of linear triangular elements
1987
Abstract We study a simple superconvergent scheme which recovers the gradient when solving a second-order elliptic problem in the plane by the usual linear elements. The recovered gradient globally approximates the true gradient even by one order of accuracy higher in the L 2 -norm than the piecewise constant gradient of the Ritz—Galerkin solution. A superconvergent approximation to the boundary flux is presented as well.
A new method for optimal synthesis of wavelet-based neural networks suitable for identification purposes
1999
Abstract This paper deals with a new method for optimal synthesis of Wavelet-Based Neural Networks (WBNN) suitable for identification purposes. The method uses a genetic algorithm (GA) combined with a steepest descent technique and least square techniques for both optimal selection of the structure of the WBNN and its training. The method is applied for designing a predictor for a chaotic temporal series
Estimation of Granger causality through Artificial Neural Networks: applications to physiological systems and chaotic electronic oscillators
2021
One of the most challenging problems in the study of complex dynamical systems is to find the statistical interdependencies among the system components. Granger causality (GC) represents one of the most employed approaches, based on modeling the system dynamics with a linear vector autoregressive (VAR) model and on evaluating the information flow between two processes in terms of prediction error variances. In its most advanced setting, GC analysis is performed through a state-space (SS) representation of the VAR model that allows to compute both conditional and unconditional forms of GC by solving only one regression problem. While this problem is typically solved through Ordinary Least Sq…
Learning spatial filters for multispectral image segmentation.
2010
International audience; We present a novel filtering method for multispectral satel- lite image classification. The proposed method learns a set of spatial filters that maximize class separability of binary support vector machine (SVM) through a gradient descent approach. Regularization issues are discussed in detail and a Frobenius-norm regularization is proposed to efficiently exclude uninformative filters coefficients. Experiments car- ried out on multiclass one-against-all classification and tar- get detection show the capabilities of the learned spatial fil- ters.
Unitary Space–Time Constellation Design Based on the Chernoff Bound of the Pairwise Error Probability
2008
Unitary space-time constellation design is considered for noncoherent multiple-antenna communications, where neither the transmitter nor the receiver knows the fading coefficients of the channel. By employing the Clarke's subdifferential theorem of the sum of the kappa largest singular values of a unitary matrix, we present a numerical optimization procedure for finding unitary space-time signal constellations of any dimension. The Chernoff bound of the pairwise error probability is used directly as a design criterion. The constellations are found by performing gradient descent search on a family ldquosurrogaterdquo functions that converge to the maximum pairwise error probability. The comp…
What is the best fitting function? Evaluation of lactate curves with common methods from the literature
2015
Using the lactate threshold for training prescription is the gold-standard, although there are several open questions. One open question is: What is the best fitting method for the load-lactate data points? This investigation re-analyses over 3500 lactate diagnostic datasets in swimming. Our evaluation software examines six different fitting methods with two different minimization criteria (RMSE and SE). Optimization of parameters of the functions is put in excecution with gradient descent. From a mathematical point of view, the double phase model, which consists of two linear regression lines, shows the least errors (RMSE min 0.254 ± 0.172; SE min 0.311 ± 0.210). However, this method canno…
Weight Adaptation Stability of Linear and Higher-Order Neural Units for Prediction Applications
2018
This paper is focused on weight adaptation stability analysis of static and dynamic neural units for prediction applications. The aim of this paper is to provide verifiable conditions in which the weight system is stable during sample-by-sample adaptation. The paper presents a novel approach toward stability of linear and higher-order neural units. A study of utilization of linear and higher-order neural units with the foundations on stability of the gradient descent algorithm for static and dynamic models is addressed.
A Region-based Gauss-Newton Approach to Real-Time Monocular Multiple Object Tracking
2018
We propose an algorithm for real-time 6DOF pose tracking of rigid 3D objects using a monocular RGB camera. The key idea is to derive a region-based cost function using temporally consistent local color histograms. While such region-based cost functions are commonly optimized using first-order gradient descent techniques, we systematically derive a Gauss-Newton optimization scheme which gives rise to drastically faster convergence and highly accurate and robust tracking performance. We furthermore propose a novel complex dataset dedicated for the task of monocular object pose tracking and make it publicly available to the community. To our knowledge, it is the first to address the common and…
Online shortest paths with confidence intervals for routing in a time varying random network
2018
International audience; The increase in the world's population and rising standards of living is leading to an ever-increasing number of vehicles on the roads, and with it ever-increasing difficulties in traffic management. This traffic management in transport networks can be clearly optimized by using information and communication technologies referred as Intelligent Transport Systems (ITS). This management problem is usually reformulated as finding the shortest path in a time varying random graph. In this article, an online shortest path computation using stochastic gradient descent is proposed. This routing algorithm for ITS traffic management is based on the online Frank-Wolfe approach.…
Learning the relevant image features with multiple kernels
2009
This paper proposes to learn the relevant features of remote sensing images for automatic spatio-spectral classification with the automatic optimization of multiple kernels. The method consists of building dedicated kernels for different sets of bands, contextual or textural features. The optimal linear combination of kernels is optimized through gradient descent on the support vector machine (SVM) objective function. Since a na¨ive implementation is computationally demanding, we propose an efficient model selection procedure based on kernel alignment. The result is a weight — learned from the data — for each kernel where both relevant and meaningless image features emerge after training. E…