2,979 publications from this institution
Feature selection is a useful preprocessing strategy when dealing with the classification and interpretation of high-dimensional biomedical data, especially when the sample size is small. A classification technique, exploiting parallelization efficiencies, is presented where a set of multi-layer perceptrons are trained on randomly selected feature subsets with varying cardinality. This technique is tested using high-dimensional biomedical spectra acquired from a magnetic resonance spectrometer. The classification results are benchmarked against a conventional multi-layer perceptron architecture as well as linear discriminant analysis. The new technique had a significantly lower classification error than either of the benchmarks.
In this paper, we introduce a new category of logic neurons- unineurons that are based on the concept of uninorms. As uninorms form a certain generalization of the generic categories of fuzzy set operators such as t-norms and t-conorms, the proposed unineurons inherit their logic processing capabilities which make them flexible and logically appealing. We discuss several fundamental categories of uninorms (such as UNI_or, UNI_and, and alike). In particular, we focus on the interpretability of networks composed of unineurons leading to several categories of rules to be exploited in rule-based systems. The learning aspects of the unineurons are presented along with detailed optimization schemes. Experimental results tackle two categories of problems such as: (a) a logic approximation of fuzzy sets, and (b) a design of associations between information granules where the ensuing development schemes directly relate to the fundamentals of granular (fuzzy) modeling
Developing a reasonable and efficient emergency material scheduling plan is of great significance to decreasing casualties and property losses. Real-world emergency material scheduling (EMS) problems are typically large-scale and possess complex constraints. An evolutionary algorithm (EA) is one of the effective methods for solving EMS problems. However, the existing EAs still face great challenges when dealing with large-scale EMS problems or EMS problems with equality constraints. To handle the above challenges, we apply the idea of a variable reduction strategy (VRS) to an EMS problem, which can accelerate the optimization process of the used EAs and obtain better solutions by simplifying the corresponding EMS problems. Firstly, we define an emergency material allocation and route scheduling model, and a variable neighborhood search and NSGA-II hybrid algorithm (VNS-NSGAII) is designed to solve the model. Secondly, we utilize VRS to simplify the proposed EMS model to enable a lower dimension and fewer equality constraints. Furthermore, we integrate VRS with VNS-NSGAII to solve the reduced EMS model. To prove the effectiveness of VRS on VNS-NSAGII, we construct two test cases, where one case is based on a multi-depot vehicle routing problem and the other case is combined with the initial 5∙12 Wenchuan earthquake emergency material support situation. Experimental results show that VRS can improve the performance of the standard VNS-NSGAII, enabling better optimization efficiency and a higher-quality solution.
A characteristic approach of approximate reasoning is the partial matching of observations to prototypes. This analysis is cast in the framework of fuzzy set theory and brings another dimension to the fuzzy matching criterion; this dimension is the measure of uncertainty through the concept of subjective entropy. While a similarity measure, in the matching process, activates relevant prototypes, the entropy formalism derived provides a measure of uncertainty in the partial matching to each prototype. The novel approach formalizes an entropy weights activation of prototypes for fuzzy partial matching. A methodology is developed for matching of observation to a set of prototypes making use of a suitable aggregation done with a framework of fuzzy integrals. A method of dealing with compound hypothesis is also developed.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">></ETX>
This paper is concerned with the use of radial basis function (RBF) neural networks aimed at an approximation of nonlinear mappings from R(n) to R. The study is devoted to the design of these networks, especially their layer composed of RBF's, using the techniques of fuzzy clustering. Proposed is an idea of conditional clustering whose main objective is to develop clusters (receptive fields) preserving homogeneity of the clustered patterns with regard to their similarity in the input space as well as their respective values assumed in the output space. The detailed clustering algorithm is accompanied by extensive simulation studies.
Fuzzy multimodeling is concerned with the design and utilization of families rather than a single model. The intent is to approximate data that are originated by phenomena whose nature is more relation based than function oriented. In general, fuzzy multimodels comprise a collection of local models M/sub 1/, M/sub 2/,/spl middot//spl middot//spl middot/,M/sub C/ along with the relevant mechanisms of their triggering and aggregating, aimed at assuring a suitable interaction between these models. The idea of multimodeling is contrasted with some other approaches to fuzzy modeling available in the current literature. The algorithmic details are laid down and illustrated through several detailed simulation studies.
The concept of forty information becomes a cornerstone of processing and handling linguistic data. As opposed to numeric information whose processing is well known and fully supported by a vast number of algorithms by entering the area of linguistic information processing we are immediately faced with a genuine need to revisit the fundamental concepts. We first review a notion of information granularity as a primordial concept playing a key role in human cognition. Dwelling on that, the study embarks on the concept of communication with fuzzy sets. In particular, we discuss a so-called fuzzy communication channel. The ideas of communication exploiting forty information call for its efficient encoding and decoding that subsequently leads to minimal losses of transmitted information. Interestingly enough, the incurred losses depend heavily on the granularity of the linguistic information involved-in this way one can take advantage of the level of uncertainty residing within the transmitted information granules.
In this study, we introduce a concept of granular worlds and elaborate on various representation and communication issues arising therein. A granular world embodies a collection of information granules being regarded as generic conceptual entities used to represent knowledge and handle problem solving. Granular computing is a paradigm supporting knowledge representation, coping with complexity, and facilitating interpretation of processing. In this sense, it is crucial to all man-machine pursuits and data mining and intelligent data analysis, in particular. There are two essential facets that are inherently associated with any granular world, that is a formalism used to describe and manipulate information granules and the granularity of the granules themselves (roughly speaking, by the granularity we mean a “size” of such information granules; its detailed definition depends upon the formal setting of the granular world). There are numerous formal models of granular worlds ranging from set-theoretic developments (including sets, fuzzy sets, and rough sets) to probabilistic counterparts (random sets, random variables and alike). In light of the evident diversity of granular world (occurring both in terms of the underlying formal settings as well as levels of granularity), we elaborate on their possible interaction and identify implications of such communication. More specifically, we have cast these in the form of the interoperability problem that is associated with the representation of information granules. © 2000 John Wiley & Sons, Inc.
As the geometry thickness error of composite radome impacts the electromagnetic (EM) performance of the antenna-radome system, a novel interval arithmetic analytic approach to the analysis of the effect on average power pattern of antenna-radome system with the thickness error in the composite radome-based is proposed. The thickness error of radome's composite material is modeled as interval-valued errors. The link between the interval of thickness error and the interval power pattern along with some main EM characteristics (sidelobe level, peak power, and half-power beamwidth) expressed as intervals are efficiently constructed. Some comparisons with measured and simulated results reported in the state-of-the-art literature, experiment data and Monte Carlo (MC) method result serve as a way to validate the interval analysis (IA)-based method. Some numerical examples are reported to reveal the effect of the main geometric properties (i.e., location, size, width, and midpoint of the error interval) of the thickness error interval on power pattern. The obtained results show that the proposed IA-based approach offers tangible advantages and exhibits effectiveness versus some traditional statistical techniques (e.g., MC method).
This paper introduces an information granularity reduction principle in connection with the analysis of the component of uncertainty associated with data. This overall study is illustrated utilizing simple numerical studies dealing with dynamical systems with first order dynamics. Classical and fuzzy Petri models are introduced in the analysis of dynamical systems. The overall study is illustrated utilizing simple numeric studies. The agenda involves a number of essential development issues: (i) providing a constructive way to build Petri nets out of numerical experimental data from dynamical systems, (ii) analyzing the component of uncertainty associated with data and elaborating on its minimization via an optimal quantization of the variables involved in the model of construction, (iii) considering the role of set-theoretic and fuzzy set frameworks in the transformation of numeric quantities into their qualitative (symbolic) counterparts, and (iv) identifying the role of Petri nets in the analysis of dynamical systems.
