Catálogo de publicaciones - libros

In this paper we propose a new approach to data mining and knowledge discovery based on information flow distribution in a flow graph. Flow graphs introduced in this paper are different from those proposed by Ford and Fulkerson for optimal flow analysis and they model flow distribution in a network rather than the optimal flow which is used for information flow examination in decision algorithms. It is revealed that flow in a flow graph is governed by Bayes’ rule, but the rule has an entirely deterministic interpretation without referring to its probabilistic roots. Besides, a decision algorithm induced by a flow graph and dependency between conditions and decisions of decision rules is introduced and studied, which is used next to simplify decision algorithms.

This article gives an overview of the Rough Set Exploration System (RSES). RSES is a freely available software system toolset for data exploration, classification support and knowledge discovery. The main functionalities of this software system are presented along with a brief explanation of the algorithmic methods used by RSES. Many of the RSES methods have originated from rough set theory introduced by Zdzisław Pawlak during the early 1980s.

From the granular computing perspective, the existing notions of validity, confidence, and coverage of rules in approximation spaces may be viewed as too crisp since granularity of the space is not, in general, taken into account in their definitions. In this article, an extension of the classical approach to a general rough case is discussed. We introduce and investigate graded validity, confidence, and coverage of rules as examples of rough validity, confidence, and coverage, respectively. The graded notions are based on the concepts of graded meaning of formulas and sets of formulas, studied in our earlier works. Among others, the notions of graded validitity, confidence, and coverage refine and extend the classical forms by taking into account granules of information drawn toward objects of an approximation space.

Most substations today contain a large number of Intelligent Electronic Devices (IEDs), each of which captures and stores locally measured analogue signals, and monitors the operating status of plant items. A key issue for substation data analysis is the adequacy of our knowledge available to describe certain concepts of power system states. It may happen sometimes that these concepts cannot be classified crisply based on the data/information collected in a substation. The paper therefore describes a relatively new theory based on rough sets to overcome the problem of overwhelming events received at a substation that cannot be crisply defined and for detecting superfluous, conflicting, irrelevant and unnecessary data generated by microprocessor IEDs. It identifies the most significant and meaningful data patterns and presents this concise information to a network or regionally based analysis system for decision support. The operators or engineers can make use of the summary of report to operate and maintain the power system within an appropriate time. The analysis is based on time-dependent event datasets generated from a PSCAD/EMTDC simulation. A 132/11 kV substation network has been simulated and various tests have been performed with a realistic number of variables being logged to evaluate the algorithms.

This article presents experiments aiming at testing the effectiveness of the implemented low-level descriptors for automatic recognition of musical instruments and musical styles. The paper discusses first some problems in audio information analysis related to MPEG-7-based applications. A short overview of the MPEG-7 standard focused on audio information description is also given. System assumptions for automatic identification of music and musical instrument sounds are presented. A discussion on the influence of descriptor selection process on the classification accuracy is included. Experiments are carried out basing on a decision system employing Rough Sets (RS) and Artificial Neural Networks (ANNs).

Computational intelligence poses several possibilities in Bioinformatics, particularly by generating low-cost, low-precision, good solutions. Rough sets promise to open up an important dimension in this direction. The present article surveys the role of artificial neural networks, fuzzy sets and genetic algorithms, with particular emphasis on rough sets, in Bioinformatics. Since the work entails processing huge amounts of incomplete or ambiguous biological data, the knowledge reduction capability of rough sets, learning ability of neural networks, uncertainty handling capacity of fuzzy sets and searching potential of genetic algorithms are synergistically utilized.

This article introduces an ethological approach to evaluating biologically-inspired collective behavior in intelligent systems. This is made possible by considering ethology (ways to explain agent behavior) in the context of approximation spaces. The aims and methods of ethology in the study of the behavior of biological organisms were introduced by Niko Tinbergen in 1963. The rough set approach introduced by Zdzisław Pawlak provides a ground for concluding to what degree a particular behavior for an intelligent system is a part of a set of behaviors representing a norm or standard. A rough set approach to ethology in studying the behavior of cooperating agents is introduced. Approximation spaces are used to derive action-based reference rewards for a swarm. Three different approaches to projecting rewards are considered as a part of a study of learning in real-time by a swarm. The contribution of this article is the introduction of an approach to rewarding swarm behavior in the context of an approximation space.

In this paper, we discuss approximation spaces in a granular computing framework. Such approximation spaces generalise the approaches to concept approximation existing in rough set theory. Approximation spaces are constructed as higher level information granules and are obtained as the result of complex modelling. We present illustrative examples of modelling approximation spaces that include approximation spaces for function approximation, inducing concept approximation, and some other information granule approximations. In modelling of such approximation spaces we use an important assumption that not only objects but also more complex information granules involved in approximations are perceived using only partial information about them.

The paper describes the “Rough Sets Database System” (called in short the RSDS system) for the creation of a bibliography on rough sets and their applications. This database is the most comprehensive online rough sets bibliography currently available and is accessible from the RSDS website at . This service has been developed to facilitate the creation of a rough sets bibliography for various types of publications. At the moment the bibliography contains over 1900 entries from more than 815 authors. It is possible to create the bibliography in or format. In order to broaden the service contents it is possible to append new data using a specially dedicated online form. After appending data online the database is updated automatically. If one prefers sending a data file to the database administrator, please be aware that the database is updated once a month. In the present version of the RSDS system, we have broadened information about the authors as well as the Statistics sections, which facilitates precise statistical analysis of the service. In order to widen the abilities of the RSDS system we added new features including:

We present a novel approach to understanding the concepts of the theory of rough sets in terms of the inverse probabilities derivable from data. It is related to the Bayes factor known from the Bayesian hypothesis testing methods. The proposed Rough Bayesian model (RB) does not require information about the prior and posterior probabilities in case they are not provided in a confirmable way. We discuss RB with respect to its correspondence to the original Rough Set model (RS) introduced by Pawlak and Variable Precision Rough Set model (VPRS) introduced by Ziarko. We pay a special attention on RB’s capability to deal with multi-decision problems. We also propose a method for distributed data storage relevant to computational needs of our approach.