We propose a classification of measures enabling to compare fuzzy characterizations of objects, according to their properties and the purpose of their utilization. We establish

Boosting is recognized as one of the most successful techniques for generating classifier ensembles. Typically, the classifier outputs are combined by the weighted majority vote. The purpose of this study is to demonstrate the advantages of some fuzzy combination methods for ensembles of classifiers designed by Boosting. We ran 2-fold cross-validation experiments on 6 benchmark data sets to compare the fuzzy and non-fuzzy combination methods. On the "fuzzy side" we used the fuzzy integral and the decision templates with different similarity measures. On the "non-fuzzy side" we tried simple combiners such as the majority vote, minimum, maximum, average, product, and the Naive Bayes combination. Surprisingly, the minimum, maximum, average and product, which have been reported elsewhere to work very well on a variety of problems, appeared to be inadequate for our task. Thus the real contest was among the fuzzy combination methods on the one hand, and the weighted majority vote, the simple majority vote, and the Naive Bayes combiner, on the other hand. In our experiments, the fuzzy methods performed consistently better than the nonfuzzy methods. The weighted majority vote showed a stable performance, though slightly inferior to the performance of the fuzzy combiners. The majority vote and the Naive Bayes combiners had erratic behavior, ranging from the best to the worst contestants for different data sets.

A typical content-based image retrieval (CBIR) system would need to handle the vagueness in the user queries as well as the inherent uncertainty in image representation, similarity measure, and relevance feedback. In this paper, we discuss how fuzzy set theory can be effectively used for this purpose and describe an image retrieval system called FIRST (Fuzzy Image Retrieval SysTem) which incorporates many of these ideas. FIRST can handle exemplar-based, graphical-sketch-based, as well as linguistic queries involving region labels, attributes, and spatial relations. FIRST uses Fuzzy Attributed Relational Graphs (FARGs) to represent images, where each node in the graph represents an image region and each edge represents a relation between two regions. The given query is converted to a FARG, and a low-complexity fuzzy graph matching algorithm is used to compare the query graph with the FARGs in the database. The use of an indexing scheme based on a leader clustering algorithm avoids an exhaustive search of the FARG database. We quantify the retrieval performance of the system in terms of several standard measures.

: Fuzzy relation-based models for handling uncertainty (in a non-probabilistic way) in diagnosis problems have been introduced by Sanchez about twenty years ago, and since then applied and developed by several researchers. The paper first reviews the existing proposals, also including the association-based abductive model proposed by Reggia and his co-workers, which uses non-fuzzy relations. A new model is presented in order to have a more expressive representation capable of handling uncertainty, and also of distinguishing, i) between manifestations which are certainly absent from those which are not (yet) observed, and ii) between manifestations which cannot be caused by a given disorder and manifestations for which we do not know if they can or cannot be caused by this disorder. This new model is based on possibility theory and the so-called twofold fuzzy sets, previously introduced by the authors. Key words : Fuzzy relation equation ; diagnosis; abduction ; possibility ; certainty....

Abstract. Several researchers on country-of-origin (coo) have expressed their interest in knowing how consumers ’ emotional reactions toward coo-cues affect product attitude formation. This paper shows how Fuzzy Set Theory might serve as a useful approach to that problem. Data was gathered by means of selfadministered questionnaires. Technically, orness of OWA-operators enabled us to distinguish consumers expressing highly positive versus less positive emotions toward coo. It appeared that this variance in emotional estate goes together with a difference in aggregating product-attribute beliefs. 1 Introduction to Product Attitude Formation We start this paper by providing a short overview of the literature on product attitude formation. Without going into all the details, it provides the larger (marketing) context in which the technical contribution of our paper should be seen. The motivation for this is that the contribution of this paper does not only lay in technical aspects of

Usually, decisions are based on a number of criteria, leading to the intuitive idea that solutions scoring better on all criteria should outperform solutions with lower scores on the criteria. This feature is not respected in current classification algorithms that learn by examples, making them unsuitable for decision analysis. This work probes via a semantical approach into the fundamentals of ranking solutions in such a supervised learning context, and proposes a framework (elementary and stochastic). This results in a novel instancebased algorithm (Ordinal Stochastic Dominance Learner) and an adaption of classification trees ( Ranking Trees)..

We present in this article an improvement of the operator triples π of Yager and Rybalov: it is a new operator called mean 3π. Whereas triple π of Yager is an operator completely reinforced this new operator is a mean operator, which makes it more robust to the noise

Abstract — In this paper we propose a method for generating OWA weighting vectors from the individual assessments on a set of alternatives in such a way that these weights maximize the consensus among individual assessments with respect to the outcome provided by the OWA operator.

Most information retrieval systems based on linguistic approaches use symmetrically and uniformly distributed linguistic term sets to express the weights of queries and the relevance degrees of documents. However, to improve the system-user interaction it seems more adequate to express these linguistic weights and degrees by means of unbalanced linguistic scales, i.e., linguistic term sets with different discrimination levels on both sides of mid linguistic term. In this contribution we present an information retrieval system which accepts weighted queries whose weights are expressed using unbalanced linguistic term sets. Then, system provides the retrieved documents classified in linguistic relevance classes assessed on unbalanced linguistic term sets. To do so, we propose a methodology to manage unbalanced linguistic information and we use the linguistic 2-tuple model as representation base of the unbalanced linguistic information. Additionally, the linguistic 2-tuple model allows us to increase the number of relevance classes in the output and also to improve the performance of information retrieval system. Index Terms Information retrieval systems, weighted query, unbalanced linguistic term set, computing with words. I.

Abstract not found