TL;DR: This work provides two algorithms that extend the celebrated junction tree algorithm, process soft evidence, and have different efficiency characteristics, and provides methodological guidance to model soft evidence in the form of beliefs on single and multiple variables, propositional logical formulae, and even conditional distributions.

TL;DR: An approach to dealing with missing data, both during the design and normal operation of a neuro-fuzzy classifier is presented, and a number of simulation results for well-known data sets are provided in order to illustrate the properties and performance of the proposed approach.

TL;DR: It is shown how the definitions generalize themselves, and that the equivalence between conditional non-interactivity and conditional doxastic independence still has the equivalent in the marginal case.

TL;DR: It is concluded that when the algorithm starts near the optimum, UMDAc is able to reach it, and the speed of convergence to the optimum decreases as the dimension increases.

TL;DR: It is proved that doxastic independence and non-interactivity are equivalent in the context of the transferable belief model.

TL;DR: An approximate algorithm is presented to obtain a posteriori intervals of probability, when available information is also given with intervals, using probability trees as a means of representing and computing with the convex sets of probabilities associated to the intervals.

TL;DR: This paper proposes a model for the parametric representation of linguistic hedges in Zadeh?s fuzzy logic that yields a method of efficiently computing linguistic truth expressions accompanied with a rich algebraic structure of the linguistic truth domain, namely De Morgan algebra.

TL;DR: It is proved that convergence of the search distribution to the global optima for the factorized distribution algorithm (FDA) if thesearch distribution is a Boltzmann distribution and the size of the population is large enough.

TL;DR: Several types of temporal noisy gates are introduced, which constitute a generalization of traditional canonical models of multicausal interactions, such as the noisy OR-gate, which have been usually applied to static domains.

TL;DR: The paper employs a heuristic function to reduce the search space for finding the solution of the classical problem of computing approximate max–min inverse fuzzy relation, an NP-complete problem for which no polynomial time algorithm is known.

TL;DR: This paper demonstrates how a priori knowledge of parameter dependencies, even incomplete knowledge, can be incorporated to efficiently obtain accurate models that account for parameter interdependencies.

TL;DR: A new separation criterion called L-separation is introduced and its main properties are studied and it is shown how it allows to represent the above-mentioned independence models through directed acyclic graphs.

TL;DR: By employing the proposed FLC, the closed-loop system performance can be designed and it does not require the complex process of finding a common Lyapunov function for a large number of fuzzy sub-systems in order to guarantee the system stability.

TL;DR: An iterative approximation procedure is implemented based on state-space abstraction methods for computing approximate probabilities with Bayesian networks by aggregating the states of variables and demonstrates the desirable anytime property in experiments.

TL;DR: A decision-theoretic method is developed that yields approximate, low cost troubleshooting plans by making more relevant observations and devoting more time to generate a plan, which is robust with respect to changes in observation and repair costs.

TL;DR: This special issue is focused on cross-fertilization aspects between probabilistic graphical models and evolutionary computation, and a recently proposed metaheuristic named ant colony optimization is used to learn Bayesian network structures from data.

TL;DR: In this article, a measure of contradiction between fuzzy rules is defined and a minimum degree of consistency exists in the rule base, and a process of attenuation is carried out between the rules that do not comply with this degree.

TL;DR: A new approximation method is proposed based on a new concept of incomplete belief potentials that allows to compute simultaneously lower and upper bounds for belief and plausibility and can be used for a resource-bounded propagation scheme.

TL;DR: This paper proposes a new algorithm for evolutionary multi-objective optimization by learning and using probabilistic mixture distributions, which uses a specialized diversity preserving selection operator and is named MIDEA.

TL;DR: This work shows how to use the matrix notation for several problems related to belief functions and the transferable belief model.

TL;DR: This paper generalizes the well-known exponential and linear convex T–S aggregation operators into a wider class of compensatory aggregation operators, built as the composition of an arbitrary quasi-linear mean with a t-norm and at-conorm, which are called quasi- linear T– S operators.

TL;DR: This paper proposes a new algorithm for learning BNs based on a recently introduced metaheuristic, which has been successfully applied to solve a variety of combinatorial optimization problems: ant colony optimization (ACO).

TL;DR: A novel approach is presented to fine tune a direct fuzzy controller based on very limited information on the nonlinear plant to be controlled through a two-stage algorithm.

TL;DR: This method may be seen as a hierarchical clustering procedure applied to the columns of a binary data matrix, using a particular dissimilarity measure to compute approximations of the mass functions, which can be combined efficiently in the coarsened frame using the fast Mobius transform algorithm.

TL;DR: Fuzzy models, in which features are partitioned into linguistic terms, will be compared to probabilistic rule-based models with the same number of terms in every linguistic partition and it is proposed to use these probabilism models to estimate a lower precision limit which fuzzy rule learning algorithms should surpass.

TL;DR: Sufficient conditions are derived for robust asymptotic output tracking controllers in the format of linear matrix inequalities (LMIs), which can be very efficiently solved by using LMI optimization techniques.

TL;DR: It is shown that the use of interval computations can simplify the (rather complicated) proofs of the subjective probabilities p 1 and p 2.

TL;DR: The paper analyzes the applicability of the methods for learning Bayesian networks in the context of genetic and evolutionary search and concludes that the combination of the two approaches yields robust, efficient, and accurate search.

TL;DR: The links between the logical and the graphical frameworks in both numerical and quantitative settings are dealt with and a translation of these graphs into possibilistic bases is provided.