Theory of T-norms and fuzzy inference methods

In this paper, the theory of T-norm and T-conorm is reviewed and the T-norm, T-conorm and negation function are defined as a set of T-operators. Some typical T-operators and their mathematical properties are presented. Finally, the T-operators are extended to the conventional fuzzy reasoning methods which are based on the and operators. This extended fuzzy reasoning provides both a general and a flexible method for the design of fuzzy logic controllers and, more generally, for the modelling of any decision-making process.     

基于神经网络的模糊推理

为了使模糊推理符合推理原则，目前已定义了１０多种模糊关系，但各种模糊关系定义都存在一定的缺陷。本文提出的基于神经网络的模糊推理，能很好地符合模糊推理原则。     

Happy and sad thoughts: An exploration of children's integer reasoning

The purpose of this study was to investigate elementary children's conceptions that might serve as foundations for integer reasoning. Working from an abstract algebraic perspective and using an opposite-magnitudes context that is relevant to children, we analyzed the reasoning of 33 children in grades K-5. We focus our report on three prominent ways of reasoning. We do this by describing and analyzing the responses of three particular children (in Grades 1, 3, and 5) who exemplify these ways of reasoning. We view each of the three ways of reasoning as rich and interesting, and we see relationships of each to formal integer reasoning. At the same time, we view these ways of reasoning in terms of increasing levels of sophistication, potentially belonging to a single learning trajectory. Thus, we see the roots of more sophisticated integer reasoning in children's early intuitions about opposite magnitudes.     

Symmetric implicational method of fuzzy reasoning

Fuzzy reasoning should take into account the factors of both the logic system and the reasoning model, thus a new fuzzy reasoning method called the symmetric implicational method is proposed, which contains the full implication inference method as its particular case. The previous full implication inference principles are improved, and unified forms of the new method are respectively established for FMP (fuzzy modus ponens) and FMT (fuzzy modus tollens) to let different fuzzy implications be used under the same way. Furthermore, reversibility properties of the new method are analyzed from some conditions that many fuzzy implications satisfy, and it is found that its reversibility properties seem fine. Lastly, the more general α-symmetric implicational method is put forward, and its unified forms are achieved.     

Triple I method of approximate reasoning on Atanassov's intuitionistic fuzzy sets

Two basic inference models of fuzzy reasoning are fuzzy modus ponens (FMP) and fuzzy modus tollens (FMT). The Triple I method is a very important method to solve the problems of FMP and FMT. The aim of this paper is to extend the Triple I method of approximate reasoning on Atanassov's intuitionistic fuzzy sets. In the paper, we first investigate the algebra operators' properties on the lattice structure of intuitionistic fuzzy information and provide the unified form of residual implications which indicates the relationship between intuitionistic fuzzy implications and fuzzy implications. Then we present the intuitionistic fuzzy reasoning version of the Triple I principles based on the models of intuitionistic fuzzy modus ponens (IFMP) and intuitionistic fuzzy modus tollens (IFMT) and give the Triple I method of intuitionistic fuzzy reasoning for residual implications. Moreover, we discuss the reductivity of the Triple I methods for IFMP and IFMT. Finally, we propose α-Triple I method of intuitionistic fuzzy reasoning.     

Fuzzy sets in approximate reasoning,Part 1: Inference with possibility distributions

A survey of about twenty years of approximate reasoning based on fuzzy logic and possibility theory is proposed. It is not only made as an annotated bibliography of past works. It also emphasizes simple basic ideas that govern most of the existing methods, especially the principle of minimum specificify and the combination/projection principle that facilitate a comparison between fuzzy set-based methods and other numerical approaches to automated reasoning. Also, a significant part of the text is devoted to the representation of truth-qualified, certainty-qualified and possibility-qualified fuzzy statements. A new attempt to classify the numerous models of fuzzy “if … then” rules from a semantic point of view is presented. In the past, people have classified them according to algebraic properties of the underlying implication, or by putting constraints on the expected behavior of the inference process (by analogy with classical logic), or by running extensive comparative trials of particular implications on test-examples. Here the classification is based on whether the rules qualify the truth, the certainty or the possibility of their conclusions. Each case corresponds to a specific way of deriving the underlying conditional possibility distribution. This paper focuses on semantic approaches to approximate reasoning based on fuzzy sets, commonly exemplified by the generalized modus ponens, but also considers applications to current topics in Artificial Intelligence such as default reasoning and qualitative process modeling. A companion survey paper is devoted to syntax-oriented methods.     

基于三角范数的模糊逻辑中的三Ⅰ方法

三I推理方法是一种新的模糊推理方法,通过已有的研究成果表明,在许多方面它优于传统的CRI推理方法,它将成为模糊系统和人工智能的理论和应用研究中一个比较理想的推理机制。最近,国外学者提出了一个新的模糊逻辑形式系统,叫做Monoidal t-norm based logics(简记为MTL),已经证明这个形式系统是所有基于左连续三角范数的模糊逻辑的共同形式化。本文基于这类逻辑将三I推理方法形式化,从而在这些逻辑系统中为三推理方法找到了可靠的逻辑依据。     

A fuzzy approach to R&D project portfolio selection

A major advance in the development of project selection tools came with the application of options reasoning in the field of Research and Development (R&D). The options approach to project evaluation seeks to correct the deficiencies of traditional methods of valuation through the recognition that managerial flexibility can bring significant value to projects. Our main concern is how to deal with non-statistical imprecision we encounter when judging or estimating future cash flows. In this paper, we develop a methodology for valuing options on R&D projects, when future cash flows are estimated by trapezoidal fuzzy numbers. In particular, we present a fuzzy mixed integer programming model for the R&D optimal portfolio selection problem, and discuss how our methodology can be used to build decision support tools for optimal R&D project selection in a corporate environment.     

模糊推理的摄动性

在模糊推理过程中,推理前件的微小变化往往会引起推理结果的较大变化。针对这种情况,本文讨论基本模糊推理,多重模糊推理以及多重多维模糊推理的摄动性问题,并根据推理过程中复合算子和蕴涵算子的不同选取对各种形式的模糊推理的最大摄动参数进行评估。     

基于极大模糊熵原理的模糊推理三I算法

用模糊熵来度量模糊推理结果的模糊程度，并用本文给出的极大模糊熵原理对王国俊先生提出的模糊推理三Ⅰ算法做进一步的解释和改进，提出基于极大模糊熵原理的模糊推理三Ⅰ算法，证明这几种算法在一定条件下是关系再现算法。     

Fuzzy criteria for feature selection

The presence of less relevant or highly correlated features often decrease classification accuracy. Feature selection in which most informative variables are selected for model generation is an important step in data-driven modeling. In feature selection, one often tries to satisfy multiple criteria such as feature discriminating power, model performance or subset cardinality. Therefore, a multi-objective formulation of the feature selection problem is more appropriate. In this paper, we propose to use fuzzy criteria in feature selection by using a fuzzy decision making framework. This formulation allows for a more flexible definition of the goals in feature selection, and avoids the problem of weighting different goals is classical multi-objective optimization. The optimization problem is solved using an ant colony optimization algorithm proposed in our previous work. We illustrate the added value of the approach by applying our proposed fuzzy feature selection algorithm to eight benchmark problems.     

Toward extended fuzzy logic—A first step

Fuzzy logic adds to bivalent logic an important capability—a capability to reason precisely with imperfect information. Imperfect information is information which in one or more respects is imprecise, uncertain, incomplete, unreliable, vague or partially true. In fuzzy logic, results of reasoning are expected to be provably valid, or p-valid for short. Extended fuzzy logic adds an equally important capability—a capability to reason imprecisely with imperfect information. This capability comes into play when precise reasoning is infeasible, excessively costly or unneeded. In extended fuzzy logic, p-validity of results is desirable but not required. What is admissible is a mode of reasoning which is fuzzily valid, or f-valid for short. Actually, much of everyday human reasoning is f-valid reasoning.f-Valid reasoning falls within the province of what may be called unprecisiated fuzzy logic, FLu. FLu is the logic which underlies what is referred to as f-geometry. In f-geometry, geometric figures are drawn by hand with a spray pen—a miniaturized spray can. In Euclidean geometry, a crisp concept, C, corresponds to a fuzzy concept, f-C, in f-geometry. f-C is referred to as an f-transform of C, with C serving as the prototype of f-C. f-C may be interpreted as the result of execution of the instructions: Draw C by hand with a spray pen. Thus, in f-geometry we have f-points, f-lines, f-triangles, f-circles, etc. In addition, we have f-transforms of higher-level concepts: f-parallel, f-similar, f-axiom, f-definition, f-theorem, etc. In f-geometry, p-valid reasoning does not apply. Basically, f-geometry may be viewed as an f-transform of Euclidean geometry.What is important to note is that f-valid reasoning based on a realistic model may be more useful than p-valid reasoning based on an unrealistic model.     

A compensation-based recurrent fuzzy neural network for dynamic system identification

In this paper, a type of compensation-based recurrent fuzzy neural network (CRFNN) for identifying dynamic systems is proposed. The proposed CRFNN uses a compensation-based fuzzy reasoning method, and has feedback connections added in the rule layer of the CRFNN. The compensation-based fuzzy reasoning method can make the fuzzy logic system more adaptive and effective, and the additional feedback connections can solve temporal problems. The CRFNN model is proven to be a universal approximator in this paper. Moreover, an online learning algorithm is proposed to automatically construct the CRFNN. The results from simulations of identifying dynamic systems have shown that the convergence speed of the proposed method is faster than the convergence speed of conventional methods and that only a small number of tuning parameters are required.     

一种基于模糊推理的因素权重确定方法

在模糊综合评价中,如何确定因素的权重是非常重要的,也是非常困难的,它直接关系到评价结果的合理性与真实性.利用模糊推理技术,提出了一种基于模糊推理的因素权重确定方法,其利用人们处理复杂问题的逻辑思维方式,建立评价空间和单位评价空间的映射,得到模糊关系方程并用于确定因素的权重.实验结果表明,利用该模型得到的权重能反映各因素的重要性,用其对待评对象进行模糊综合,其结果也是合理的.     

Relating and extending semantical approaches to possibilistic reasoning

Two main semantical approaches to possibilistic reasoning with classical propositions have been proposed in the literature. Namely, Dubois-Prade's approach known as possibilistic logic, whose semantics is based on a preference ordering in the set of possible worlds, and Ruspini's approach that we redefine and call similarity logic, which relies on the notion of similarity or resemblance between worlds. In this article we put into relation both approaches, and it is shown that the monotonic fragment of possibilistic logic can be semantically embedded into similarity logic. Furthermore, to extend possibilistic reasoning to deal with fuzzy propositions, a semantical reasoning framework, called fuzzy truth-valued logic, is also introduced and proved to capture the semantics of both possibilistic and similarity logics.     

On the suitability of minimum and product operators for the intersection of fuzzy sets

In many papers concerning fuzzy set theory it is assumed that the membership or an element in the intersection of two or more fuzzy sets is given by the minimum of product of the corresponding membership values. To use these operators in modelling aspects of the real world, such as decision making, however, it is necessary to prove their appropriateness empirically. The main question of this study is whether people rating the membership of objects in the intersection of two fuzzy sets behave in accordance with one of these models. An important problem in answering this question is how to measure membership which seems to have the characteristics of an absolute scale. No measurement structure is available at present, but a practical method for scaling is suggested. The results of our experiments indicate that neither the product nor the minimum fit the data sufficiently well, but the latter seems to be preferable.     

Arc refractor methods for adaptive importance sampling on large Bayesian networks under evidential reasoning

Approximate Bayesian inference by importance sampling derives probabilistic statements from a Bayesian network, an essential part of evidential reasoning with the network and an important aspect of many Bayesian methods. A critical problem in importance sampling on Bayesian networks is the selection of a good importance function to sample a network’s prior and posterior probability distribution. The initially optimal importance functions eventually start deviating from the optimal function when sampling a network’s posterior distribution given evidence, even when adaptive methods are used that adjust an importance function to the evidence by learning. In this article we propose a new family of Refractor Importance Sampling (RIS) algorithms for adaptive importance sampling under evidential reasoning. RIS applies “arc refractors” to a Bayesian network by adding new arcs and refining the conditional probability tables. The goal of RIS is to optimize the importance function for the posterior distribution and reduce the error variance of sampling. Our experimental results show a significant improvement of RIS over state-of-the-art adaptive importance sampling algorithms.     

The Logic of Analogy

How should a scientist argue when the data are insufficient to allow him to reason by classical or statistical models? After all, in most real world situations - in business or in war - that is the unhappy norm. In such cases the ordinary man instinctively argues by analogy, as Leibniz long ago showed; indeed if time presses, there is no alternative. The trouble, however, is that if we then include such arguments in our scientific reasoning, then, as we all know, this can lead to false conclusions. To escape from this dilemma, is there any alternative logical basis from which we can start our reasoning? What is proposed here is that instead of the well tried three valued logic of true, false or probable, we should adopt the three valued logic of true, false or possible. A rational system for analogue arguments can then be developed by these means, and with it the advantages brought by the use of symbols and so on. Such a method, however, includes many necessary changes as to how to structure our problems and how to apply new criteria; and it is some of these changes that are outlined in this note. For instance, it outlines the meaning of ‘causal relationships’ in analogue arguments, as well as how to define ‘rational choice’ in terms of analogue propositions. The advantage throughout is that this allows us to argue with less rather than more data.     

On students’ conceptions of arithmetic average: the case of inference from a fixed total

There is more to understanding the concept of mean than simply knowing and applying the add-them-up and divide algorithm. In the following, we discuss a component of understanding the mean – inference from a fixed total – that has been largely ignored by researchers studying students understanding of mean. We add this component to the list of types of reasoning needed to understand mean and discuss student responses to tasks designed to elicit this component of reasoning. These responses reveal that inference from a fixed total reasoning is rare even in advance high school students.