J. Vandewalle

• KU Leuven

Networks represent patterns of interactions between components of complex systems present in nature, science, technology and society. Furthermore, graph theory allows to perform insightful analysis for different kinds of data by representing the instances as nodes of a weighted network, where the weights characterize similarity between the data poi...

Recently, the easy-to-implement state-dependent Riccati equation (SDRE) strategy has been shown effective for numerous practical applications. Since being similar to SDRE, the newly emerged state-dependent differential Riccati equation (SDDRE) approach shares most of the benefits of SDRE, and exhibits interesting potential from both the analytical...

This paper presents the development of a multi campus, multidisciplinary curriculum for graduated engineers. Focus is placed on the societal role of engineering and the potential added value of technology for organisations in the social profit sector and for vulnerable groups in society that are the target public of these social profit organisation...

The development of two educational workshops, one on energy efficiency and one on human-machine interfaces, is detailed and discussed. Attraction to engineering is not created as much as lost at early ages through current education methods. Through positive, hands-on experiences with engineering in K-12 education, this trend can be turned. IEEE stu...

To investigate the role of the CYP17 gene promoter polymorphism in the pathobiology of uterine leiomyomas in African and Caucasian women. During a 6-month period, 145 Caucasian and black South African women undergoing hysterectomy were included prospectively. Blood samples were obtained for DNA analysis. Factors modifying the risk for uterine leiom...

Support Vector Machines Basic Methods of Least Squares Support Vector Machines Bayesian Inference for LS-SVM Models Robustness Large Scale Problems LS-SVM for Unsupervised Learning LS-SVM for Recurrent Networks and Control.

In this paper a master-slave synchronization scheme for Lur'e systems is investigated, which consists of vector field modulation by means of the message signal.

Support vector machines (SVM's) have been introduced in literature as a method for pattern recognition and function estimation, within the framework of statistical learning theory and structural risk minimization. A least squares version (LSSVM) has been recently reported which expresses the training in terms of solving a set of linear equations in...

This report is available by anonymous ftp from ftp.esat.kuleuven.ac.be in the directory pub/SISTA/delathauwer/reports/ldl-95-45.ps.Z ESAT - Katholieke Universiteit Leuven, Kardinaal Mercierlaan 94, 3001 Leuven (Heverlee), Belgium, tel 32/16/32 18 05, fax 32/16/32 19 86, email: Lieven.DeLathauwer@esat.kuleuven.ac.be. Lieven De Lathauwer is a researc...

It is investigated how a given higher-order tensor can be approximated in least-squares sense by a rank-1 tensor. A multilinear equivalent of the power method is proposed. It is shown that the tensor approximation yields a new line of approach to the problem of Independent Component Analysis.

This dissertation gives a number of contributions to the field of controller reduction, by providing some new frequency weighted or time-domain weighted model reduction procedures, and showing equivalences of certain reduction algorithms. 1. It is proven that Enns's well-known reduction scheme Frequency Weighted Balanced Truncation (FWBT) with cert...

We develop a technique for Blind Source Separation based on simultaneous diagonalization of (linear com-binations of) third-order tensor \slices" of the fourth-order cumulant. It will be shown that, in a Jacobi-type iteration scheme, the computation of an elementary ro-tation can be reformulated in terms of a simultaneous matrix diagonalization.

In this Chapter we treat the problem of nonlinear system identification using neural networks. Model structures and their parametrization by multilayer perceptrons are discussed, together with learning algorithms, practical aspects and examples. The Chapter is organized as follows. In Section 3.1 we review model structures such as NARX, NARMAX and...

In this Chapter we provide the reader with some background material on neural control strategies and we discuss neural optimal control in more detail. The Chapter is organized as follows. In Section 4.1 the basic principles of existing methods in neural control are presented, including direct and indirect adaptive control, reinforcement learning, n...

In this Chapter we develop a model based neural control framework which consists of neural state space models and neural state space controllers. Like in modern (robust) control theory standard plant forms are considered. In order to analyse and synthesize neural controllers within this framework, the so-called NLq system form is introduced. NLq s...

In this Chapter we discuss two basic types of artificial neural network architectures that are used in the sequel for modelling and control purposes: the multilayer perceptron and the radial basis function network. This Chapter is organized as follows. In Section 2.1 we give a description of the architectures. In Section 2.2 we present an overview...

In this book we discussed the use of artificial neural networks for modelling and control of nonlinear systems in a systemtheoretical context. After a short introduction on neural information processing systems in Chapter 1, we have reviewed basic neural network architectures and their learning rules in Chapter 2, for feedforward as well as recurre...

The research on the parameter estimation of a sum of K exponentially damped sinusoids has led to the development of many estimation algorithms. In some applications, however, it is desired to prefilter the input data in order to reduce the noise and enhance the parameters of interest. This chapter presents a prefiltering technique in which a filter...

This paper addresses the problem of identifying nonlinear discrete-time multivariable systems using radial-basis-function neural networks. A recursive identification scheme is proposed that first builds recursively an RBF (Radial-Basis-Function) neural net model structure and then uses a stable recursive weight updating algorithm to modify the weig...

An on-line nonlinear system identification scheme is proposed based on the idea of neural nets structure adaptation. Using the so-called RBF (Radial-Basis-Function) neural nets as generic model structure, we have been able to derive a stable and efficient approach including the structural generation, grid adaptation and the weight update. Main conv...

Recently severed subspace methods have appeared in the literature for multivariable discrete-time state space identification, where state space models are computed directly from input/output data. These state space identification methods are viewed as the better alternatives to polynomial model identification, owing to the better numerical conditio...

A geometrically inspired solution strategy is presented to three fundamental linear algebraic problems that have many applications in system theory and network analysis and simulution. Problem 1: Given an n multiplied by 1 vector p. Find all negative n multiplied by 1 vectors x that are orthogonal to p: p**t multiplied by X equals 0. Problem 2: Giv...

Total linear least squares (TLLS) is a method of solving over-determined sets of linear equations Ax equals b where both the observation vector b and the datamatrix A are inaccurate. The technique has been introduced by Golub and Van Loan and amounts to fitting a best subspace to left bracket A, b right bracket . The concept and geometric interpret...

In the case of multiple poles, the classical method of partial fraction expansion (PFE), so often used for the computation of the inverse Z-transform of a rational function, leads to cumbersome expressions for the obtained discrete-time sequences. Exploiting the degrees of freedom left in the PFE, it is possible to obtain time sequences of a simple...

In the case of multiple poles, the classical method of partial fraction expansion (PFE), so often used for the computation of the inverse Z-transform of a rational function, leads to cumbersome expressions for the obtained discrete-time sequences. Exploiting the degrees of freedom left in the PFE, it is possible to obtain time sequences of a simple...

Adapting Silverman's structure algorithm a general, efficient and recursive algorithm is given for the following problem. Given the Laurent expansion at α of a non-singular rational matrix A(p), compute the Laurent expansion of the inverse A(p)−1 at α. The use of this algorithm in the inversion of rational and polynomial matrices and in particular...

The research on the parameter estimation of a sum of K exponentially damped sinusoids has led to the development of many estimation algorithms. In some applications, however, it is desired to prefilter the input data in order to reduce the noise and enhance the parameters of interest. In this paper, we present a prefiltering technique in which a fi...