Search

This paper deals with the evaluation problem of Japanese traditional crafts in which product items are evaluated according so-called kansei features by means of the semantic differential method. As for traditional crafts, decisions on which items to buy or use are usually influenced by personal feelings/characteristics, we shall propose a consumer-oriented evaluation model targeting on those recommendation requests specified by consumers. Particularly, given a consumer's recommendation request, the proposed model aims at defining an evaluation function that quantifies how well a product item meets the consumer's feeling preference. An application for evaluating patterns of Kutani porcelain is conducted to illustrate how the proposed evaluation model works practically. [ABSTRACT FROM AUTHOR]

We treat the sentences in a finite inconsistent knowledge base as assertions that are true with probability at least some primary threshold η and consider as consequences those assertions entailed to have probability at least some secondary threshold ζ. [ABSTRACT FROM AUTHOR]

Uncertainty is the main obstacle in predicting crude oil price. Although there are various models and computational methods on crude oil price forecasting in literature, most of them do not effectively predict the variability of crude oil price due to uncertainty. Very recently, Hu and He [2] reported of using ILS (Interval Least Square) approach to forecast the stock market and obtained much better results than that obtained with traditional point methods, In this paper, we investigate if the ILS approach can forecast the relationship between commodity inventory levels and crude oil spot prices effectively. Our empirical study suggests that both the ILS method and the confidence interval method can produce comparable quality forecasts. While the computational result produced by the ILS method seems slightly worse than the 95% confidence intervals in two quality measurements, the differences are negligible. On a new forecasting quality measurement proposed in this paper, the ILS method produces results better than the 95% confidence intervals. Hence, interval method is a feasible alternative in crude oil price forecasting. [ABSTRACT FROM AUTHOR]

Decision making of the severeness of the acute radiation syndrome is a major challenge due to the fact that the radiation dose is not known and cannot easily be reconstructed. Although radiation accidents are relatively rare, an automatic damage classification for individual medical prognosis of the health status of a patient is necessary. Under the threat of nuclear terroristic attacks the problem has received special attention. Early classification of the severeness of the damage in case of the acute radiation syndrome allows for separation of patients with irreversibly damaged cell renewal systems from those with reversible damage. As a consequence, the available resources for the treatment of patients can be used most efficiently and, especially, therapeutic actions for patients with irreversible cell damage can be taken in time. This paper concentrates on the damage to the hemopoietic system. Measurements from available patient data are represented by time series of cell counts of the relevant blood cell lines. Features extracted from reduced dynamic models form the basis for automatic damage classification where emphasis is on an early classification based on a time horizon not exceeding 10 days. Our newest results guarantee a generalization rate in the range of 80%. [ABSTRACT FROM AUTHOR]

This paper proposes a common framework for various probabilistic logics. It consists of a set of uncertain premises with probabilities attached to them. This raises the question of the strength of a conclusion, but without imposing a particular semantics, no general solution is possible. The paper discusses several possible semantics by looking at it from the perspective of probabilistic argumentation. [ABSTRACT FROM AUTHOR]

In this paper, we propose a unified formulation of deduction, induction and abduction using granularity based on variable precision rough set models proposed by Ziarko [12] and measure-based semantics for modal logics proposed by Murai et al. [4]. [ABSTRACT FROM AUTHOR]

Uncertainties in the real world often appear as variabilities of observed data under similar conditions. In this paper, we use interval functions to model uncertainty and function volatility. To estimate such kinds of functions, we propose a practical interval function approximation algorithm. Applying this algorithm, we have studied stock market forecasting with real economic data from 1930-2004. The computational results indicate that interval function approximation can produce better quality forecasts than that obtained with other methods. [ABSTRACT FROM AUTHOR]

In this paper, we first apply a technique of fuzzy correspondence analysis to deal with fuzziness in subjective evaluation data of traditional crafts in which evaluated objects are assessed by multiple evaluators against multiple kansei features, where each kansei feature corresponds to two (adjectival) kansei words having opposite meanings. Basically, the variation in subjective evaluation data is transformed into a parameter space trying to preserve information as much as possible, and thereby we define fuzzy sets which express relative fuzziness between principal component scores in the parameter space. Then we define a so-called fitness measure which expresses how well an object fits with a given kansei word, taking both the distance and the relative fuzziness into account. This fitness measure is then utilized together with ordered weighted averaging (OWA, for short) operator for aggregation to define a ranking function that quantifies how well an object fits with the feeling preference specified as a set of kansei words by a potential consumer. A prototyping system for visualizing the fuzzy correspondence analysis of subjective evaluation data as well as for personalized recommendation is also developed and illustrated. [ABSTRACT FROM AUTHOR]

Many types of fuzzy truth values have been proposed, such as numerical truth values, interval truth values, triangular truth values and trapezoid truth values and so on. Recently, different type of fuzzy truth values (which we call multi-interval truth values) have been proposed and discussed. A characteristic feature of multi-interval truth values is that some of the truth values are not convex. This fact makes us difficult to understand algebraic properties on multi-interval truth values. This paper first shows that a set of multi-interval truth values is de Morgan bi-lattices when the conventional logical operations are introduced. Next, this paper focuses on functions over the set of the simplest multi-interval truth values, i.e., we define a multi-interval truth value as a non-empty subset of {0, 1, 2}. Then, this paper discusses mathematical properties of functions over multi-interval truth values. [ABSTRACT FROM AUTHOR]

Type-2 fuzzy sets have come to play an increasingly important role in both applications and in the general theory of fuzzy sets. The basis of type-2 fuzzy sets is a certain algebra of truth values, as set forth by Zadeh. This paper is a survey of results about this algebra, along with some new material. [ABSTRACT FROM AUTHOR]

Ordered algebraic structures are encountered in many areas of mathematics. One frequently wishes to embed a given ordered algebraic structure into a complete ordered algebraic structure in a manner that preserves some aspects of the algebraic and order theoretic properties of the original. It is the purpose here to survey some recent results in this area. [ABSTRACT FROM AUTHOR]

In this paper we discuss the nature of independence of sources in the theory of evidence from an algebraic point of view, starting from the analogy with the case of projective geometries. Independence in Dempster's rule is equivalent to independence of frames as Boolean algebras. Collection of frames, though, can be interpreted as semi-modular lattices on which independence can be defined in several different forms. We prove those forms to be distinct but related to Boolean independence, as a step towards a more general definition of this fundamental notion. [ABSTRACT FROM AUTHOR]

In this paper we introduce the relative belief of singletons as a novel Bayesian approximation of a belief function. We discuss its nature in terms of degrees of belief under several different angles, and its applicability to different classes of belief functions. [ABSTRACT FROM AUTHOR]

In this paper we develop a rough set-based solution to dealing with the inconsistent decision classes of instances in inducing decision rules and an evidential reasoning method to resolve such inconsistent conclusions encountered in determining class decisions for instances. The distinguishing aspects of our method are to exploit the rough boundary region in inducing rules and to aggregate multiple conclusions in classifying instances. We present our proposed method and use an example to illustrate how our method can be applied to classification problems along with its advantage. [ABSTRACT FROM AUTHOR]

The main aim of this brief note is to explain relations between the classic approach to set approximations and recent proposals appearing in the literature on rough sets. In particular, relations between the standard topological concepts and basic concepts of rough set theory are considered. [ABSTRACT FROM AUTHOR]

In this paper, we investigate reducts preserving a structure induced from dominance-based rough sets as well as a structure from variable-precision dominance-based rough sets. It is shown that three kinds of reducts are obtained in the dominance-based rough set approach. For each kind of reduct, a discernibility matrix is given for the enumeration of the reducts. Moreover, it is shown that four kinds of reducts are obtained in the variable-precision dominance-based rough set approach. A counter-example showing the impossibility of reduct enumeration in the case of variable-precision dominance-based rough set approach. [ABSTRACT FROM AUTHOR]

Although fuzzy set and probability theories are complementary for dealing with pervasive imprecision and uncertainty in real world problems, object-oriented database models combining the relevance and strength of both the theories appear to be sporadic. This paper introduces our extension of Eiter et al.'s probabilistic object base model with two key features: (1) uncertain and imprecise attribute values are represented as probability distributions on a set of fuzzy set values; and (2) class methods with uncertain and imprecise input and output arguments are formally integrated into the new model. A probabilistic interpretation of relations on fuzzy set values is proposed for their combination with probability degrees. Then the syntax and semantics of fuzzy-probabilistic object base schemas, instances, and selection operation are defined. Furthermore, the soft computing paradigm needs to have real systems implemented to be useful in practice. This paper also presents our development of FPDB4O as a management system for fuzzy and probabilistic object bases of the proposed model. [ABSTRACT FROM AUTHOR]

Many multiple attribute decision analysis (MADA) problems are characterised by both quantitative and qualitative attributes with various types of uncertainties. Incompleteness (or ignorance) and vagueness (or fuzziness) are among the most common uncertainties in decision analysis. The evidential reasoning (ER) and the interval grade evidential reasoning (IER) approaches have been developed in recent years to support the solution of MADA problems with interval uncertainties and local ignorance in decision analysis. In this paper, the ER approach is enhanced to deal with both interval uncertainty and fuzzy beliefs in assessing alternatives on an attribute. In this newly developed FIER approach, local ignorance and grade fuzziness are modelled under the integrated framework of a distributed fuzzy belief structure, leading to a fuzzy belief decision matrix. A numerical example is provided to illustrate the detailed implementation process of the FIER approach and its validity and applicability. [ABSTRACT FROM AUTHOR]

The authors previous work on probabilistic constraint reasoning assumes the uncertainty of numerical variables within given bounds, characterized by a priori probability distributions. It propagates such knowledge through a network of constraints, reducing the uncertainty and providing a posteriori probability distributions. An inverse problem aims at estimating parameters from observed data, based on some underlying theory about a system behavior. This paper describes how nonlinear inverse problems can be cast into the probabilistic constraint framework, highlighting its ability to deal with all the uncertainty aspects of such problems. [ABSTRACT FROM AUTHOR]

Among modern applications of document retrieval, a practical system for retrieving scientific publications has recently been attracting much attention from research community. In a scientific document, there are many types of uncertainty information occurring, such as research areas of the documents or authors. Thus, a method for efficiently handling uncertainty information when retrieving scientific information, as well as other kinds of uncertainty information, is currently desirable. In our paper, we propose a novel fuzzy retrieval framework based on approximating reasoning for document retrieval. We also discuss using approximating reasoning to discover additional relations in the database to support more advanced search functions in an intelligent information retrieval system. This paper also introduces an experimental system implementing our proposed technique. The performance of the system is then evaluated and analyzed. [ABSTRACT FROM AUTHOR]

An alternative perspective on granular modelling is introduced where an information granule characterises the relationship between a label expression and elements in an underlying perceptual space. Label semantics is proposed as a framework for representing information granules of this kind. Mass relations and linguistic decision trees are then introduced as two types of granular models in label semantics. Finally, its shown how linguistic decision trees can be combined within an attribute hierarchy to model complex multi-level composite mappings. [ABSTRACT FROM AUTHOR]

The main goal of information fusion can be seen as improving human or automatic decision-making by exploiting diversities in information from multiple sources. High-level information fusion aims specifically at decision support regarding situations, often expressed as "achieving situation awareness". A crucial issue for decision making based on such support is trust that can be defined as "accepted dependence", where dependence or dependability is an overall term for many other concepts, e.g., reliability. This position paper reports on ongoing and planned research concerning imprecise probability as an approach to improved dependability in high-level information fusion. We elaborate on high-level information fusion from a generic perspective and a partial mapping from a taxonomy of dependability to high-level information fusion is presented. Three application domains: defense, manufacturing, and precision agriculture, where experiments are planned to be implemented are depicted. We conclude that high-level information fusion as an application-oriented research area, where precise probability (Bayesian theory) is commonly adopted, provides an excellent evaluation ground for imprecise probability. [ABSTRACT FROM AUTHOR]

Support Vector Machines (SVM) is one of the most widely used technique in machines leaning. After the SVM algorithms process the data and produce some classification, it is desirable to learn how well this classification fits the data. There exist several measures of fit, among them the most widely used is kernel target alignment. These measures, however, assume that the data are known exactly. In reality, whether the data points come from measurements or from expert estimates, they are only known with uncertainty. As a result, even if we know that the classification perfectly fits the nominal data, this same classification can be a bad fit for the actual values (which are somewhat different from the nominal ones). In this paper, we show how to take this uncertainty into account when estimating the quality of the resulting classification. [ABSTRACT FROM AUTHOR]

In many practical situations, we are not satisfied with the accuracy of the existing measurements. There are two possible ways to improve the measurement accuracy: first, instead of a single measurement, we can make repeated measurements; the additional information coming from these additional measurements can improve the accuracy of the result of this series of measurements;second, we can replace the current measuring instrument with a more accurate one; correspondingly, we can use a more accurate (and more expensive) measurement procedure provided by a measuring lab - e.g., a procedure that includes the use of a higher quality reagent. In general, we can combine these two ways, and make repeated measurements with a more accurate measuring instrument. What is the appropriate trade-off between sample size and accuracy? In our previous paper, we solved this problem for the case of static measurements. In this paper, we extend the results to the case of dynamic measurements. [ABSTRACT FROM AUTHOR]

In many practical situations, we are not satisfied with the accuracy of the existing measurements. There are two possible ways to improve the measurement accuracy: first, instead of a single measurement, we can make repeated measurements; the additional information coming from these additional measurements can improve the accuracy of the result of this series of measurements;second, we can replace the current measuring instrument with a more accurate one; correspondingly, we can use a more accurate (and more expensive) measurement procedure provided by a measuring lab - e.g., a procedure that includes the use of a higher quality reagent. In general, we can combine these two ways, and make repeated measurements with a more accurate measuring instrument. What is the appropriate trade-off between sample size and accuracy? This is the general problem that we address in this paper. [ABSTRACT FROM AUTHOR]

In many areas of science and engineering, it is desirable to estimate statistical characteristics (mean, variance, covariance, etc.) under interval uncertainty. For example, we may want to use the measured values x(t) of a pollution level in a lake at different moments of time to estimate the average pollution level; however, we do not know the exact values x(t) - e.g., if one of the measurement results is 0, this simply means that the actual (unknown) value of x(t) can be anywhere between 0 and the detection limit DL. We must therefore modify the existing statistical algorithms to process such interval data. Such a modification is also necessary to process data from statistical databases, where, in order to maintain privacy, we only keep interval ranges instead of the actual numeric data (e.g., a salary range instead of the actual salary). Most resulting computational problems are NP-hard - which means, crudely speaking, that in general, no computationally efficient algorithm can solve all particular cases of the corresponding problem. In this paper, we overview practical situations in which computationally efficient algorithms exist: e.g., situations when measurements are very accurate, or when all the measurements are done with one (or few) instruments. [ABSTRACT FROM AUTHOR]

This paper is about a short survey of some basic uncertainty measures in systems analysis arising from coarse data, together with new modeling results on upper semicontinuous random processes, viewed as random fuzzy sets. Specifically, we present the most general mathematical framework for analyzing coarse data, such as random fuzzy data, which arise often in information systems. Our approach is based upon the theory of continuous lattices. This probabilistic analysis is also useful for investigating upper semicontinuous random functions in stochastic optimization problems. [ABSTRACT FROM AUTHOR]

We will give a state-of-the-art survey of the study of substructural logics. Originally, substructural logics were introduced as logics which, when formulated as Gentzen-style systems, lack some of the three basic structural rules, i.e. contraction, weakening and exchange. For example, relevance logics and linear logic lack the weakening rule, many-valued logics, fuzzy logics and linear logic lack the contraction rule, and hence all of them can be regarded as substructural logics. These logics have been studied extensively and various theories have been developed for their investigation. However their study has been carried out independently, mainly due to the different motivations behind them, avoiding comparisons between different substructural logics. [ABSTRACT FROM AUTHOR]