Guest Editorial: Fuzzy logic and soft computing in an effective and efficient processing of imperfect information - in memoriam Ashley Morris
Hierarchical text categorization using fuzzy relational thesaurus
Text categorization is the classification to assign a text document to an appropriate category in a predefined set of categories. We present a new approach for the text categorization by means of Fuzzy Relational Thesaurus (FRT). FRT is a multilevel category system that stores and maintains adaptive local dictionary for each category. The goal of our approach is twofold; to develop a reliable text categorization method on a certain subject domain, and to expand the initial FRT by automatically added...
Highly robust training of regularizedradial basis function networks
Radial basis function (RBF) networks represent established tools for nonlinear regression modeling with numerous applications in various fields. Because their standard training is vulnerable with respect to the presence of outliers in the data, several robust methods for RBF network training have been proposed recently. This paper is interested in robust regularized RBF networks. A robust inter-quantile version of RBF networks based on trimmed least squares is proposed here. Then, a systematic comparison...
Homogeneous aggregation operators
Recently, the utilization of invariant aggregation operators, i.e., aggregation operators not depending on a given scale of measurement was found as a very current theme. One type of invariantness of aggregation operators is the homogeneity what means that an aggregation operator is invariant with respect to multiplication by a constant. We present here a complete characterization of homogeneous aggregation operators. We discuss a relationship between homogeneity, kernel property and shift-invariance...
How to secure a high quality knowledge base in a rulebased system with uncertainty
Although the first rule-based systems were created as early as thirty years ago, this methodology of expert systems designing still proves to be useful. It becomes especially important in medical applications, while treating evidence given in an electronic format. Constructing the knowledge base of a rule-based system and, especially, of a system with uncertainty is a difficult task because of the size of this base as well as its heterogeneous character. The base consists of facts, ordinary rules...
Imposing restrictions on density functions utilised in computing with words
Applying the generalised extension principle within the area of Computing with Words typically leads to complex maximisation problems. If distributed quantities-such as, e.g., size distributions within human populations-are considered, density functions representing these distributions become involved. Very often the optimising density functions do not resemble those found in nature; for instance, an optimising density function could consist of two single Dirac pulses positioned near the opposite...
Improving the generalization ability of neuro-fuzzy systems by ε-insensitive learning
A new learning method tolerant of imprecision is introduced and used in neuro-fuzzy modelling. The proposed method makes it possible to dispose of an intrinsic inconsistency of neuro-fuzzy modelling, where zero-tolerance learning is used to obtain a fuzzy model tolerant of imprecision. This new method can be called ε-insensitive learning, where, in order to fit the fuzzy model to real data, the ε-insensitive loss function is used. ε-insensitive learning leads to a model with minimal Vapnik-Chervonenkis...
Inference in conditional probability logic
An important field of probability logic is the investigation of inference rules that propagate point probabilities or, more generally, interval probabilities from premises to conclusions. Conditional probability logic (CPL) interprets the common sense expressions of the form “if ..., then ...” by conditional probabilities and not by the probability of the material implication. An inference rule is probabilistically informative if the coherent probability interval of its conclusion is not necessarily...
Inference in expert systems based on complete multivalued logic
Influence of modeling structure in probabilistic sequential decision problems
Markov Decision Processes (MDPs) are a classical framework for stochastic sequential decision problems, based on an enumerated state space representation. More compact and structured representations have been proposed: factorization techniques use state variables representations, while decomposition techniques are based on a partition of the state space into sub-regions and take advantage of the resulting structure of the state transition graph. We use a family of probabilistic exploration-like...
Information boundedness principle in fuzzy inference process
The information boundedness principle requires that the knowledge obtained as a result of an inference process should not have more information than that contained in the consequent of the rule. From this point of view relevancy transformation operators as a generalization of implications are investigated.
Interpretability of linguistic variables: a formal account
This contribution is concerned with the interpretability of fuzzy rule-based systems. While this property is widely considered to be a crucial one in fuzzy rule-based modeling, a more detailed formal investigation of what “interpretability” actually means is not available. So far, interpretability has most often been associated with rather heuristic assumptions about shape and mutual overlapping of fuzzy membership functions. In this paper, we attempt to approach this problem from a more general...
Marginalization in models generated by compositional expressions
In the framework of models generated by compositional expressions, we solve two topical marginalization problems (namely, the single-marginal problem and the marginal-representation problem) that were solved only for the special class of the so-called “canonical expressions”. We also show that the two problems can be solved “from scratch” with preliminary symbolic computation.
Marginalization in multidimensional compositional models
Efficient computational algorithms are what made graphical Markov models so popular and successful. Similar algorithms can also be developed for computation with compositional models, which form an alternative to graphical Markov models. In this paper we present a theoretical basis as well as a scheme of an algorithm enabling computation of marginals for multidimensional distributions represented in the form of compositional models.
Marginalization like a projection.
This paper studies the problem of marginalizing convex polytopes of probabilities represented by a set of constraints. This marginalization is obtained as a special case of projection on a specific subspace. An algorithm that projects a convex polytope on any subspace has been built and the expression of the subspace, where the projection must be made for obtaining the marginalization, has been calculated.
MATHWARE – Special issue devoted to FLINS 2008 conference.
Method of least squares applied to the generalized modus ponens with interval-valued fuzzy sets.
Firstly we present a geometric interpretation of interval-valued fuzzy sets. Secondly, we apply the method of least squares to the fuzzy inference rules when working with these sets. We begin approximating the lower and upper extremes of the membership intervals to axb type functions by means of the method of least squares. Then we analyze a technique for evaluating the conclusion of the generalized modus ponens and we verify the fulfillment of Fukami and alumni axioms [9].
Minimal decision rules based on the apriori algorithm
Based on rough set theory many algorithms for rules extraction from data have been proposed. Decision rules can be obtained directly from a database. Some condition values may be unnecessary in a decision rule produced directly from the database. Such values can then be eliminated to create a more comprehensible (minimal) rule. Most of the algorithms that have been proposed to calculate minimal rules are based on rough set theory or machine learning. In our approach, in a post-processing stage,...
Mixed integer programming, general concept inclusions and fuzzy description logics.
Model of adaptation under indeterminacy
Information retrieval in information systems (IS) with large amounts of data is not only a matter of an effective IS architecture and design and technical parameters of computer technology used for operation of the IS, but also of an easy and intuitive orientation in a number of offers and information provided by the IS. Such retrievals in IS are, however, frequently carried out with indeterminate information, which requires other models of orientation in the environment of the IS.