2,979 publications from this institution
Fuzzy class label adjustment is a classification preprocessing strategy that compensates for the possible imprecision of class labels. Using training vectors, robust measures of location and dispersion are computed for each class center. Based on distances from these centers, fuzzy sets are constructed that determine the degree to which each input vector belongs to each class. These membership values are then used to adjust class labels for the training vectors. This strategy is evaluated using a multilayer perceptron and two different robust location measures for the discrimination of meteorological storm events and is shown to improve the performance of the underlying classifier.
A mechanism for realizing a modus ponens performed on the basis of the collection of "if-then" statements containing fuzzy premises is studied. Its essence relies on derivation of a relational mechanism transforming fuzzy truth from the space of antecedents to the space of consequents. Moreover, some considerations dealing with relevancy of the studied scheme of reasoning are included.
Imbalanced datasets play an important role in many fields in real applications such as medical diagnosis, business risk management, abnormal product testing and evaluation. In these cases, the minority classes are usually important. Granular computing has been developed and effectively applied to many problems especially imbalanced data classification. In this paper, we propose a new strategy to build information granulations (IGs) for each class separately and represent sub-attributes based on categorical values (including discretized values of the numerical attributes) to solve the overlapping among IGs. This strategy reduces the computational time, improves classification performance and considers high-balanced accuracy among classes. The experimental results on several datasets have demonstrated the effectiveness of our proposal.
Standard assumption of pattern recognition problem is that processed elements belong to recognized classes. However, in practice, we are often faced with elements presented to recognizers, which do not belong to such classes. For instance, paper-to-computer recognition technologies (e.g. character or music recognition technologies, both printed and handwritten) must cope with garbage elements produced at segmentation level. In this paper we distinguish between elements of desired classes and other ones. We call them native and foreign elements, respectively. The assumption that we have only native elements results in incorrect inclusion of foreign ones into desired classes. Since foreign elements are usually not known at the stage of recognizer construction, standard classification methods fail to eliminate them. In this paper we study construction of recognizers based on support vector machines and aimed on coping with foreign elements. Several tests are performed on real-world data.
In this study, we introduce a concept of hierarchical granular clustering and establish its algorithmic framework. We show that the proposed model naturally gives rise to information granules that are both of higher order and higher type, offering a compelling justification behind their emergence. In a concise way, we can capture the overall architecture of information granules as a hierarchy exhibiting conceptual layers of increasing abstraction: numeric data → information granules → information granules of type-2, order-2 → ... information granules of higher type/order. The elevated type of information granules is reflective of the visible hierarchical facet of processing and the inherent diversity of the individual locally revealed structures in data. While the concept and the methodology deliver some general settings, the detailed algorithmic aspects are discussed in detail when using fuzzy clustering realized by means of fuzzy c-means. Furthermore, for illustrative purposes, we mainly focus on interval-valued fuzzy sets and granular interval fuzzy sets arising at the higher level of the hierarchy. Higher type fuzzy sets are formed with the help of the principle of justifiable granularity. The conceptually sound hierarchy is established in a general way, which makes it equally applicable to various formalisms of representation of information granules. Experiments are reported for synthetic and publicly available datasets.
Nonlinear equations systems (NESs) are widely used in real-world problems while they are also difficult to solve due to their characteristics of nonlinearity and multiple roots. Evolutionary algorithm (EA) is one of the methods for solving NESs, given their global search capability and an ability to locate multiple roots of a NES simultaneously within one run. Currently, the majority of research on using EAs to solve NESs focuses on transformation techniques and improving the performance of the used EAs. By contrast, the problem domain knowledge of NESs is particularly investigated in this study, using which we propose to incorporate the variable reduction strategy (VRS) into EAs to solve NESs. VRS makes full use of the systems of expressing a NES and uses some variables (i.e., core variable) to represent other variables (i.e., reduced variables) through the variable relationships existing in the equation systems. It enables to reduce partial variables and equations and shrink the decision space, thereby reducing the complexity of the problem and improving the search efficiency of the EAs. To test the effectiveness of VRS in dealing with NESs, this paper integrates VRS into two existing state-of-the-art EA methods (i.e., MONES and DRJADE), respectively. Experimental results show that, with the assistance of VRS, the EA methods can significantly produce better results than the original methods and other compared methods.
Control engineering has permeated a vast territory of human endeavours. Its conceptual, methodological, and engineering facets as well as numerous realizations are present everywhere. Perhaps in many cases we are even not aware of the ingenuity of control solutions that are brought into everyday life. Among many books and research monographs in the control area that are available on a market today, this book authored by an expert in modern control engineering, Professor Zdzislaw Bubnicki, is unique in several different and important ways. First, it profoundly reflects upon the broad spectrum of applications of control engineering going beyond classic control (that has been predominantly focused on control of physical systems) and ventures into the ideas of control of systems when a human factor plays a pivotal role. This is particularly relevant when dealing with various control problems in the area of management, logistics, intelligent systems, and decision making. First, the book brings a wealth of such interesting and important control ideas dealing among others with the problems of task and resource distribution, assembly processes, and allocation in systems with transport factor. This holistic view at control engineering becomes a leitmotiv of the entire book. Second, it emphasizes the relevance and omnipresence of uncertainty and stresses its eminent visibility in the practice of control engineering. Thirdly, it offers an interesting vision of the paradigms of control in application to intelligent systems and architectures that have emerged in the realm of Computational Intelligence. Here, the book relates explicitly to neural networks and fuzzy systems. The author has coined a concept of uncertain variable, which offers an interesting and unique insight into a way of handling uncertainty. The reader may wish to consult 1 for more pertinent details. In a nutshell, uncertain variables build a highly coherent and unified view at various formalisms of uncertainty representation and processing including probability and fuzzy sets. There is definitely a clearly visible trend of recognition and appreciation of the existence of various facets of uncertainty. Here uncertain variables accentuate this standpoint. While the probabilistic form of uncertainty comes with a long history and associates with a broad spectrum of algorithmic pursuits, fuzzy sets play a pivotal role in all cases when human judgment comes into a picture. This important facet of a multitude of uncertainty is well reflected in the book. The chapters on relational description of uncertainty (Chapter 6), probabilistic handling of uncertainty (Chapter 7), and subsequently the chapter on fuzzy variables (Chapter 9) provide a lucid coverage of these important topics. The book comprises 13 chapters and given their content, it splits into five parts. The first part composed of two first chapters, serves as a comprehensive introduction that covers all necessary prerequisites and makes the book self-contained to some extent. The second part, composed of Chapters 3–5, is concerned with control realized for deterministic plants where no uncertainty is taken into consideration. Chapters 6–9 form the core of the book and bring a wealth of ideas when the factors of uncertainty manifest in various ways. The discussion embraces a relational and probabilistic description of uncertainty, presents uncertain variables and covers an in-depth topic of fuzzy (soft) variables. A significant portion of these chapters is devoted to some comparative studies in which the author clarifies differences between several fundamental ways of dealing with uncertainty. This looks like an excellent addition to the main stream of discussion given the fact that quite often we may encounter some conflicting opinions or superficial and groundless comments on the subject of uncertainty. Part four (including Chapters 10 and 11) is devoted to control of closed-loop systems and covers a study on stability tools applied to continuous and discrete systems (including a series of specific techniques such as e.g., describing functions). Finally, in the two last chapters, various issues of intelligent and complex control systems as well as control of systems encountered in operations research and industrial engineering are discussed. The term of intelligent systems comes here with a well-delineated semantics and deals predominantly with neural networks, logic-algebraic methods, and mechanisms of logic in knowledge representation. The exposure of the material is highly systematic and the flow of the main ideas is coherent, easy to follow, and prudently organized. The writing is lucid and concise. Perhaps some numeric illustrative examples could have added extra value for those readers who wish to see some tangible experimental evidence of the performance of the individual algorithms. Nevertheless, the detailed description of the methods could easily compensate for this otherwise very minor shortcoming. The monograph would be definitely of significant interest to researchers, practitioners, and graduate students. The strong algorithmic flavour would appeal to all those interested in the applied side of the area. The general framework in which the control engineering principles are exposed will also appeal to those interested in the application of recent control school of thought to decision making, industrial engineering, and management. All in all, this is a highly welcome and timely research monograph being a convincing and highly visible testimony to the recent trends in the methodology and practice of modern control engineering.
The performance enhancement of system identification of various plastic materials to effectively recycle the waste plastics arises as a key issue studied here. For black plastics, which contain carbon black, one is unable to discriminate it from other materials. To facilitate the identification process, Fourier transform-infrared with attenuated total reflectance is used to carry out qualitative as well as quantitative analysis of black plastics. Since a spectrum obtained in this manner constitutes highly dimensional data, feature reduction becomes necessary to extract sound features and reduce the dimensionality of the original spectrum. In this study, three types of feature extraction techniques are considered: peak detection technique, feature extraction based on the chemical characteristics, and fuzzy transform-based feature extraction to determine sound discriminative features. In order to enhance classification process, fuzzy radial basis function neural networks classifier is constructed; these architectures of the classifiers take advantage of the hybrid technologies. Based upon experimental studies, it is shown that the proposed classification system with the feature extraction techniques exhibits superior performance over the performance reported for the already studied classifiers.
