Catálogo de publicaciones - libros

Recently, the utilization of invariant aggregation operators, i.e., aggregation operators not depending on a given scale of measurement was found as a very current theme. One type of invariantness of aggregation operators is the homogeneity what means that an aggregation operator is invariant with respect to multiplication by a constant. We present here a complete characterization of homogeneous aggregation operators. We discuss a relationship between homogeneity, kernel property and shift-invariance of aggregation operators. Several examples are included.

Copulas with given diagonal have been studied in [ 4 , 10 ]. In [ 2 , 5 , 11 ] smallest and greatest (quasi-)copulas with given diagonal are constructed. Both (two-dimensional) copulas and quasi-copulas are special cases of binary 1-Lipschitz aggregation operators [ 3 , 8 ], and in [ 7 ] 1-Lipschitz aggregation operators with given diagonal (and the consequences for (quasi-)copulas) are investigated. We give constructions for smallest and greatest 1-Lipschitz aggregation operators with given opposite diagonal, allowing us to obtain most results for (quasi-)copulas with given opposite diagonal as special cases.

This paper studies some relationships between fuzzy relations, fuzzy graphs and fuzzy measure. It is shown that a fundamental theorem of Discrete Convex Analysis is derived from the theory of fuzzy measures and the Choquet integral.

Modular artificial neural networks (MANN) have been used in the last years as clasification/forecasting machine, showing improved generalization capabilities that outperform those of single networks when the search space is stratified. Time Series data could be generated by many unknown and different sources and Modular Neural Networks, in particular Mixture of Experts models, are suitable for this time series where each expert is more capable to model some region in the input space and a gating network makes an intelligent selection of the expert that will model the specific pattern. Stochastical models for time series analysis are global models limited by the requirement of stationarity of the time series and normality and independence of the residuals. However, for most real world time series present behaviors such as heteroscedasticity, sudden burst of activity, or outliers. Such data are very common in finance, insurance, seismology and so on. In this paper we propose MANN models capable of dynamically adapt their architecture to non-stationary time series when the data is generated from several sources and is affected by the presence of outliers. Simulation results based on benchmark data sets are presented to support the proposed technique.

A feedforward neural network based on multi-valued neurons is considered in the paper. It is shown that using a traditional feedforward architecture and a high functionality multi-valued neuron, it is possible to obtain a new powerful neural network. Its learning does not require a derivative of the activation function and its functionality is higher than the functionality of traditional feedforward networks containing the same number of layers and neurons. These advantages of MLMVN are confirmed by testing using Parity n , two spirals and “sonar” benchmarks, and the Mackey-Glass time-series prediction.

For the scheduling transmission over a fading channel in wireless networks, the performance increases significantly if a specialized packet scheduler is used. The properties of this scheduler demand a learning mechanism. For this purpose, a least squares support vector machine (LS-SVM) is proposed as the learning mechanism. In the SVM methodology the number of the unknown can be infinitely dimensional. The given method is illustrated by some numerical examples.

In 1995 robot soccer was introduced with the purpose to develop intelligent, cooperative multi-robot (agents) systems. Robot soccer provides a good opportunity to test control strategies and methods of Multi-Agent-Systems. From the scientific viewpoint a soccer robot is an intelligent, autonomous agent which should carry out its task in cooperative, coordinated, and communicative way with other agents. The group behavior of agents and the behavior of a single agent should be explored. One of the single agent’s behaviors is the motion control. The desired velocity of each wheel is generated and sent to the robot comparing the desired and actual position of the robot. The mostly used motion controller today is the digital PID-controller. In this paper as a “modern”, intelligent control algorithm a neural network will be introduced and tested.

A novel fuzzy neural network, called Fuzzy Kolmogorov’s Network (FKN), is considered. The network consists of two layers of neo-fuzzy neurons (NFNs) and is linear in both the hidden and output layer parameters, so it can be trained with very fast and computationally efficient procedures. Two-level structure of the rule base helps the FKN avoid the combinatorial explosion in the number of rules, while the antecedent fuzzy sets completely cover the input hyperbox. The number of rules in the FKN depends linearly on the dimensionality of input space. The validity of theoretical results and the advantages of the FKN are confirmed by a comparison with other techniques in benchmark problems and a real-world problem of electrical load forecasting.

A combined learning algorithm for a self-organizing map (SOM) is proposed. The algorithm accelerates information processing due to the rational choice of the learning rate parameter, and can work when the number of clusters is unknown, as well as when the clusters are overlapping. This is achieved via the introduction of fuzzy inference that determines the level of membership of the classified pattern to each of the available classes. For neighborhood and membership functions, raised cosine is used. This function provides more flexibility and some new properties for the self-learning and clustering procedures.

In this paper, we propose fuzzy neural controller for a dynamic connection admission control (CAC) that supports the cell loss requirement and QoS parameters for multimedia traffic in Wireless ATM networks. Our CAC algorithms explicitly compute the bandwidth required for each class of connections based on the observed traffic statistics and the declared parameters.