Catálogo de publicaciones - libros

The use of unsupervised fuzzy learning methods produces a large number of alternative classifications. This paper presents and analyzes a series of criteria to select the most suitable of these classifications. Segmenting the clients’ portfolio is important in terms of decision-making in marketing because it allows for the discovery of hidden profiles which would not be detected with other methods and it establishes different strategies for each defined segment. In the case included, classifications have been obtained via the LAMDA algorithm. The use of these criteria reduces remarkably the search space and offers a tool to marketing experts in their decision-making.

The backpropagation neural network is one of the most widely used connectionist model, especially in the solution of real life problems. The main reasons for the popularity of this model are its conceptual simplicity and its ability to tackle a broad range of problems. But, on the other hand, this architecture shows a well known problem for dealing with nonstationary data. In this paper, a variation of feedforward neural model which uses qualitative data both for feeding the network and for back propagating the error correction is presented. The data are coded by means of a fuzzy concept of local stability.

In a fedbatch process the operational strategy can consist on controling the influent substrate concentration in the reactor, by means of the input flow manipulation. Due to the repetitive characteristic of the Sequencing Batch Reactor processes, it opens the possibility to explore the information generated in previous cycles to improve the process operation, without having on-line sensors and/or a very precise analytical model. In this work an iterative learning control strategy based on a fuzzy model is proposed. It is assumed that the measurements are analytical and only a few number of them can be obtained. So, an interpolation technique is used to improve the control performance. Simulation results for a phenol biodegradation process are presented.

In this paper we analyze two recent modelling methodologies: one based on a database preprocessing and then, the application of the fuzzy C-means to highlight useful characteristics used in target selection for direct marketing which is our first study case. The second one is based on fuzzy clustering and cubic splines in the rule consequents. Some examples are given in order to illustrate the advantages and drawbacks of these methods.

We focus on possibilistic/fuzzy optimality in the framework of mathematical programming problem with a possibilistic objective function. We observe the interaction between possibilistic objective function values. Two optimality concepts, possible and necessary optimalities are reviewed. The necessary soft optimality is investigated.

A simple definition of interval-valued probability measure is given and its implications examined. Properties are discussed which allow for the analysis of mixtures of fuzzy, possibilistic, probabilistic, cloud, and interval uncertainty utilizing interval-valued probability theory. It is shown how these properties allow for optimization under uncertainty where the uncertainty is mixed (fuzzy, possibilitic, probabilistic, clouds, and interval ). An example of this type of optimization is given illustrating the usefulness and power of the concepts.

Formulations for fuzzy and possibilistic optimization abound in the literature, but few are implemented in practice. This paper investigates the theory, semantics, and efficacy of a selection of significant fuzzy and possibilistic optimization algorithms via their application to a well-known large-scale problem: the radiation therapy planning problem. The algorithms are compared, critiqued, and organized with the following objective in mind: to guide a decision maker in the selection and implementation of a fuzzy or possibilistic optimization algorithm.

A risk-minimizing portfolio model under uncertainty is discussed. In the uncertainty model, the randomness and fuzziness are evaluated respectively by the probabilistic expectation and mean values with evaluation weights and -mean functions. The means, variances and the measurements of fuzziness for fuzzy numbers/fuzzy random variables are applied in the possibility case and the necessity case, and a risk estimation is derived from both random factors and fuzzy factors in the model. By quadratic programming approach, we derive a solution of the risk-minimizing portfolio problem. It is shown that the solution is a tangency portfolio. A numerical example is given to illustrate our idea.

A pattern tree [1] is a tree which propagates fuzzy terms using different fuzzy aggregations. Each pattern tree represents a structure for an output class in the sense that how the fuzzy terms aggregate to predict such a class. Unlike decision trees, pattern trees explicitly make use of t-norms (i.e., AND) and t-conorms (OR) to build trees, which is essential for applications requiring rules connected with t-conorms explicitly. Pattern trees can not only obtain high accuracy rates in classification applications, but also be robust to over-fitting. This paper further extends pattern trees approach by assigning certain weights to different trees, to reflect the nature that different trees may have different confidences. The concept of weighted pattern trees is important as it offers an option to trade off the complexity and performance of trees. In addition, it enhances the semantic meaning of pattern trees. The experiments on British Telecom (BT) customer satisfaction dataset show that weighted pattern trees can slightly outperform pattern trees, and both of them are slightly better than fuzzy decision trees in terms of prediction accuracy. In addition, the experiments show that (weighted) pattern trees are robust to over-fitting. Finally, a limitation of pattern trees as revealed via BT dataset analysis is discussed and the research direction is outlined.

Generally, there are two main streams of theories for studying uncertainties. One is probability theory and the other is fuzzy set theory. One of the basic ideas of fuzzy set theory is how to define and interpret membership functions. In this paper, we will study tree-structured data mining model based on a new interpretation of fuzzy theory. In this new theory, fuzzy labels will be used for modelling. The membership function is interpreted as appropriateness degrees for using labels to describe a fuzzy concept. Each fuzzy concept is modelled by a distribution on the appropriate fuzzy label sets. Previous work has shown that the new model outperforms some well-known data mining models such as Naive Bayes and Decision trees. However, the fuzzy labels used in previous works were predefined. We are interested in study the influences on the performance by using fuzzy labels with different degrees of overlapping. We test a series of UCI datasets and the results show that the performance of the model increased almost monotonically with the increase of the overlapping between fuzzy labels. For this empirical study with the LDT model, we can conclude that more fuzziness implies better performance.