Antonio Artés-Rodríguez

Generalizing CMAC architecture and training

The cerebellar model articulation controller (CMAC) is a simple and fast neural-network based on local approximations. However, its rigid structure reduces its accuracy of approximation and speed of convergence with heterogeneous inputs. In this paper, we propose a generalized CMAC (GCMAC) network that considers different degrees of generalization for each input. Its representation abilities are analyzed, and a set of local relationships that the output function must satisfy are derived. An adaptive growing method of the network is also presented. The validity of our approach and methods are shown by some simulated examples.

Support vector method for robust ARMA system identification

This paper presents a new approach to auto-regressive and moving average (ARMA) modeling based on the support vector method (SVM) for identification applications. A statistical analysis of the characteristics of the proposed method is carried out. An analytical relationship between residuals and SVM-ARMA coefficients allows the linking of the fundamentals of SVM with several classical system identification methods. Additionally, the effect of outliers can be cancelled. Application examples show the performance of SVM-ARMA algorithm when it is compared with other system identification methods.

Weighted least squares training of support vector classifiers leading to compact and adaptive schemes

An iterative block training method for support vector classifiers (SVCs) based on weighted least squares (WLS) optimization is presented. The algorithm, which minimizes structural risk in the primal space, is applicable to both linear and nonlinear machines. In some nonlinear cases, it is necessary to previously find a projection of data onto an intermediate-dimensional space by means of either principal component analysis or clustering techniques. The proposed approach yields very compact machines, the complexity reduction with respect to the SVC solution is especially notable in problems with highly overlapped classes. Furthermore, the formulation in terms of WLS minimization makes the development of adaptive SVCs straightforward, opening up new fields of application for this type of model, mainly online processing of large amounts of (static/stationary) data, as well as online update in nonstationary scenarios (adaptive solutions). The performance of this new type of algorithm is analyzed by means of several simulations.

Multi-dimensional Function Approximation and Regression Estimation

In this communication, we generalize the Support Vector Machines (SVM) for regression estimation and function approximation to multi-dimensional problems. We propose a multi-dimensional Support Vector Regressor (MSVR) that uses a cost function with a hyperspherical insensitive zone, capable of obtaining better predictions than using an SVM independently for each dimension. The resolution of the MSVR is achieved by an iterative procedure over the Karush-Kuhn-Tucker conditions. The proposed algorithm is illustrated by computers experiments.

Maximization of Mutual Information for Supervised Linear Feature Extraction

In this paper, we present a novel scheme for linear feature extraction in classification. The method is based on the maximization of the mutual information (MI) between the features extracted and the classes. The sum of the MI corresponding to each of the features is taken as an heuristic that approximates the MI of the whole output vector. Then, a component-by-component gradient-ascent method is proposed for the maximization of the MI, similar to the gradient-based entropy optimization used in independent component analysis (ICA). The simulation results show that not only is the method competitive when compared to existing supervised feature extraction methods in all cases studied, but it also remarkably outperform them when the data are characterized by strongly nonlinear boundaries between classes.

Recurrent radial basis function networks for optimal symbol-by-symbol equalization

Abstract In this paper, starting from showing that a recurrent version of a radial basis function (RBF) network can compute optimal symbol-by-symbol decisions for equalizing digital channels in digital communication systems, we present structures for non-Gaussian channel equalization and delayed decisions. To reduce the complexity of the structure, which grows exponentially with the memory of the channel (like that of Viterbi detectors), we propose some simplification option, preserving parallelism and near-optimal performance. Algorithms to apply this structures to blind equalization problems, using a novel non-decision directed learning rule to estimate the channel response, are also given. A general discussion about neural equalizers, several simulation results, and some conclusions complete the paper.

Empirical risk minimization for support vector classifiers

In this paper, we propose a general technique for solving support vector classifiers (SVCs) for an arbitrary loss function, relying on the application of an iterative reweighted least squares (IRWLS) procedure. We further show that three properties of the SVC solution can be written as conditions over the loss function. This technique allows the implementation of the empirical risk minimization (ERM) inductive principle on large margin classifiers obtaining, at the same time, very compact (in terms of number of support vectors) solutions. The improvements obtained by changing the SVC loss function are illustrated with synthetic and real data examples.

Convergence of the IRWLS Procedure to the Support Vector Machine Solution

An iterative reweighted least squares (IRWLS) procedure recently proposed is shown to converge to the support vector machine solution. The convergence to a stationary point is ensured by modifying the original IRWLS procedure.

Machine learning and data mining: strategies for hypothesis generation

Strategies for generating knowledge in medicine have included observation of associations in clinical or research settings and more recently, development of pathophysiological models based on molecular biology. Although critically important, they limit hypothesis generation to an incremental pace. Machine learning and data mining are alternative approaches to identifying new vistas to pursue, as is already evident in the literature. In concert with these analytic strategies, novel approaches to data collection can enhance the hypothesis pipeline as well. In data farming, data are obtained in an ‘organic’ way, in the sense that it is entered by patients themselves and available for harvesting. In contrast, in evidence farming (EF), it is the provider who enters medical data about individual patients. EF differs from regular electronic medical record systems because frontline providers can use it to learn from their own past experience. In addition to the possibility of generating large databases with farming approaches, it is likely that we can further harness the power of large data sets collected using either farming or more standard techniques through implementation of data-mining and machine-learning strategies. Exploiting large databases to develop new hypotheses regarding neurobiological and genetic underpinnings of psychiatric illness is useful in itself, but also affords the opportunity to identify novel mechanisms to be targeted in drug discovery and development.

SVC-based equalizer for burst TDMA transmissions

Channel equalization is a major issue in digital communications and in burst time division multiple access (TDMA) transmissions is augmented by the need of transmitting an equalizing sequence in each burst. In order to attain full use of the available bandwidth in such transmissions, short training sequences are a prerequisite. In this paper we propose to use a support vector classifier (SVC) for equalizing digital communication channels because of the good generalization properties the SVC presents for short training sequences. A linear and a nonlinear equalizer suitable for short training sequences are developed using an SVC. Instead of using quadratic programming (QP) procedures for solving the SVC, as is usually done, we present an iterative re-weighted least squares procedure for attaining the SVC solution, because it is a procedure simple to implement and allows a recursive formulation that is impossible for QP schemes. We analyze the performance of the proposed algorithm using computer simulations of general TDMA communication scenarios.