Catálogo de publicaciones - libros

Ant Colony Optimization (ACO) algorithms are based on the imitation of how ants of a colony find the shortest path between the nest and the food. This result is achieved by stigmergetic information, i.e. ants deposit a chemical substance (the pheromone) on the path they follow and their movement is guided by the amount of pheromone.

The imitation of this simple mechanism is the core of any ACO algorithm. In the present contribution we propose a new pheromone updating technique with the aim of speeding up the resulting algorithm for rendering it suited to a real-time implementation.

The ACO algorithms are very dependent on the specific application of interest. In this contribution the Vehicle Routing Problem is considered and the proposed algorithm is compared with 3 classic pheromone updating methods with respect to known benchmarks.

This paper demonstrates some issues in agent interaction on the Web, which is the center point of supporting the needs of fully-realized learning GRID in the future. Of particular importance is the conversation support., with its core element, communication protocols. We propose to construct communication protocols through learning of performatives of ACL messages based on the FIPA-ACL messages. The work involves two steps: 1) converting real conversations into a markup agent communication language and then 2) inducing communication protocols based on these set of converted conversations.

A methodology for sub-symbolic semantic encoding of words is presented. The methodology uses the standard, semantically highly-structured WordNet lexical database and the SemiDiscrete matrix Decomposition to obtain a vector representation with low memory requirements in a semantic -space. The application of the proposed algorithm over all the WordNet words would lead to a useful tool for the sub-symbolic processing of texts.

In the frame of a collaboration between Department of Information technology of the University of Milan and Stem Cells Research Institute of the DIBIT- San Raffaele, Milan, learning methods are under study following known models of the Artificial Neural Networks on human neural stem cells cultured on MEA (Multielectrode Arrays) support. The MEAs are constituted by a glass support where a set of tungsten electrodes are inserted to form a lattice structured by our group following the artificial Hopfield and Kohonen models. In such a way it is possible to electrically stimulate the neurons and to record their reaction, opening the possibility to verify learning models of the Artificial neural Networks. Neurons are stimulated with digital patterns constituted by bursts of different voltages at the input electrodes, and the electrical output generated by the neurons is analyzed with advanced methods in order to highlight organized answers by the natural neural network. The experiments performed up to now show how neurons react selectively to different patterns and show similar reactions in front of the presentation of identical or similar patterns. These results suggest the possibility of using the learning capabilities of these hybrid networks in different application fields, in particular in bionic applications.

The standard SVR formulation for real-valued function approximation on multidimensional spaces is based on the ε-insensitive loss function, where errors are considered not correlated. Due to this, local information in the feature space which can be useful to improve the prediction model is disregarded. In this paper we address this problem by defining a generalized quadratic loss where the co-occurrence of errors is weighted according to a kernel similarity measure in the feature space. We show that the resulting dual problem can be expressed as a hard margin SVR in a different feature space when the co-occurrence error matrix is invertible. We compare our approach against a standard SVR on two regression tasks. Experimental results seem to show an improvement in the performance.

This paper introduces a Flexible ICA approach to a novel blind sources separation problem. The proposed on line algorithm performs the separation after the convolutive mixing of post nonlinear convolutive mixtures. The Flexibility of the algorithm is given by the on line estimation of the score function performed by Spline Neurons. Experimental results are described to show the effectiveness of the proposed technique.

We reconsider in the Algorithmic Inference framework the accuracy of a Boolean function learnt from examples. This framework is specially suitable when the Boolean function is learnt through a Support Vector Machine, since (i) we know the number of support vectors really employed as an ancillary output of the learning procedure, and (ii) we can appreciate confidence intervals of misclassifying probability exactly in function of the cardinality of these vectors. As a result we obtain confidence intervals that are up to an order narrower than those supplied in the literature, having a slight different meaning due to the different approach they come from, but the same operational function. We numerically check the covering of these intervals.

We revisit the linear regression problem in terms of a computational learning problem whose task is to identify a confidence region for a continuous function belonging in particular to the straight lines family. Within the Algorithmic Inference framework this function is deputed to explain a relation between pairs of variables that are observed through a limited sample. Hence it is a random item within the above family and we look for a partial order relation allowing us to state a cumulative distribution function over the function specifications, hence a pair of quantiles identifying the confidence region. The regions we compute in this way is theoretically and numerically attested to contain the goal function with a given confidence. Its shape is quite different from the analogous region obtained through conventional methods as a collation of confidence intervals found for the expected value of the dependent variable as a function of the independent one.

Kernel Methods are algorithms that implicitly perform a nonlinear mapping of the input data to a high dimensional Feature Space. In this paper, we present a novel Kernel Method, for clustering problems. Unlike other popular clustering algorithms that yield piecewise linear borders among data, Kernel K-Means allows to get nonlinear separation surfaces in the data. Kernel K-Means compares better with popular clustering algorithms, on a synthetic dataset and two UCI real data benchmarks.

Bayesian Neural Networks — considering priors and averaging model results accordingly with weights probabilities - can be an important resource in solving classification problems whose learning sets have few samples. Hybrid Monte Carlo Markov Chains (HMCMC) are typically used to numerically solve the integrals involved in learning procedures; in this work a Genetic Algorithm is proposed as alternative to gradient measure to hybridize MCMC so that multimodal distribution can be better fitted and derivative calculation needed for gradient information can be omitted.