Learning fuzzy measures from data: Simplifications and optimisation strategies

Learning k-maxitive fuzzy measures from data by mixed integer programming

Failure mode and effects analysis (FMEA) is a popular and useful approach applied to examine potential failures in different products, designs, processes, and services. As a vital index, the risk priority number (RPN) can determine the risk priorities of failure modes by some risk factors such as occurrence (O), severity (S), and detection (D). However, in FMEA, the traditional risk priority number approach has some shortcomings, especially in setting the weight of risk factors. This paper presents an improved risk priority number approach based on a fuzzy measure and fuzzy integral. A fuzzy measure is used to reflect the importance of the individual indicators and the indicator set and a fuzzy integral is a nonlinear function defined on the basis of fuzzy measure. The weights of risk factors given by domain experts are seen as fuzzy densities to generate a λ -fuzzy measure which can reflect the weights’ difference and relevance about risk factors. Then, the Choquet integral is used to fuse every value of risk factors about failure modes so as to obtain the comprehensive evaluation result. The result can reflect the comprehensive risk level, so it has a definite physical significance. Finally, an illustrative example and a comparison with another approach are given to show the effectiveness of the proposed approach in the paper.

In this paper, we propose a new model for linear fractional programming problem with absolute value functions. The major contribution of this paper is that transformation of linear fractional programming problem into separate linear programming problems with some theorems and then solution of these problems by popular algorithm. This work also used to simplex type algorithms to arrive at an optimal solution for a linear programming problem with absolute value. Moreover, we compare this method with an existing method. Numerical experiments are also given to illustrate the assertions.

In biological problems such as protein sequence family identification and profile building the additive hypothesis of the probability measure is not well suited for modeling HMM based profiles because of a high degree of interdependency among homologous sequences of the same family . Fuzzy measure theory which is an extension of the classical additive theory is obtained by replacing the additive requirement of classical measures with weaker properties of monotonicity, continuity and semi-continuity. The strong correlations and the sequence preference involved in the protein structures make fuzzy measure architecture based models as suitable candidates for building profiles of a given family since fuzzy measures can handle uncertainties better than classical methods . In this paper we investigate the different measures(S-decomposable, λ and belief measures) of fuzzy measure theory for building profile models of protein sequence problems. The proposed fuzzy measure models have been tested on globin and kinase families . The results obtained from the fuzzy measure models establish the superiority of fuzzy measure theory compared to classical probability measures for biological sequence problems.

Topological analysis of intuitionistic fuzzy distance measures with applications in classification and clustering

We address the problems of multiple criteria decision-making and multiobjective optimization in which the criteria are mutually dependent. A practical application of land combat vehicle selection and its potential extension to force design for Australian Defence is considered. The criteria dependencies are modeled and formalised using the theory of fuzzy measures. The main challenge is an exponentially large set of parameters that quantify all criteria interactions. We develop a number of strategies to 1) simplify the model construction and reduce the number of parameters; 2) elicit preferences from numerical simulations and from decision makers; and 3) translate these preferences into a fuzzy measure learning problem from unlabeled data. We formulate and test a mathematical programming problem to determine fuzzy measure parameters for a relatively large number of decision criteria.

As we have shown, the simplex method refers to the so-called finite methods which allow one to find a solution for any problem of linear programming or to prove its unsolvability performing a finite number of elementary operations (addition, subtraction, multiplication, division, comparison of two real numbers). It stands to reason that the number of operations depends on the dimension (n, m) of the problem (n is the number of variables, m is the number of equality and inequality constraints). The practice of solving linear programming problems has shown that the simplex method and its modifications are very effective. It is accepted as a fact that in the majority of linear programming problems the number of elementary operations which are necessary for their solution is of the order O(n 2 m + m 2 n) [1, 91]. Here and in the sequel we denote by O(a) the quantities for which |O(a)| ≤ C(a), where C is a positive constant independent of a. We have found out that there exist “poor” linear programming problems in which the amount of elementary operations required for their solution by the simplex method is estimated by the number which exponentially depends on n and m.

This article addresses a special class of non-linear programming problems, viz., linear fractional programming problems having multiple objectives. In solving the real-life linear fractional optimization problems, the ambiguity and hesitation in the decision are inherent and ever-present, therefore, it is perfectly viable to formulate and solve these optimization models using the intuitionistic fuzzy environment. The purpose of this study is to propose a simple and computationally efficient approach to obtain the solution of multiple objective linear fractional programming problems having all the decision variables and parameters expressed in terms of triangular intuitionistic fuzzy numbers. The proposed solution algorithm is primarily based on the goal programming approach, fuzzy-based linearization technique, and a membership function strategy. The original linear fractional programming problem is first converted to its equivalent deterministic/crisp multi-objective linear fractional optimization problem using the weighted goal programming methodology along with the linear membership technique to resolve the intuitionistic fuzzy constraints into the crisp one. Finally, the variable transformation technique for the under- and over-deviational variables of the goal programming model is employed to linearize all the fractions involved in the problem so as to convert the original problem to an equivalent linear optimization problem. Further, this linear programming problem can be solved using any available commercial packages. Moreover, a numerical illustration is provided to demonstrate the steps of the proposed technique followed by the analysis and solution of an E-education set-up problem. The discussion and comparisons of the practical case establish the relevancy and usefulness of the proposed model.

This paper uses fuzzy sets and fuzzy numbers to reduce the subjectivity in the assignment of qualitative criteria in a design problem. Through fuzzy measures and the Choquet integral it is possible to combine multiple qualitative and quantitative criteria into a single numerical value and make a proper design decision. The fuzzy measures, indicators of the importance of single criteria are determined using a linear program optimization method. The input for the linear program is only a preference order of some of the alternatives. With the verified fuzzy measures it is then possible to compute a single numerical value for the remaining alternatives. This results in an order of preferred design constructions and a recommendation for an optimized design. The theory is applied to EEGIEOG (electroencephalography/electrooculography) electrode placement for sleep monitoring by incorporating the criteria: comfort, reliability and power consumption.

Using different dominance criteria in stochastic fuzzy linear multiobjective programming: A case of fuzzy weighted objective function

In this paper, we propose a method of solving the fully fuzzy linear fractional programming problems, Express all the parameters and variables are triangular fuzzy numbers. Convert all the triangular fuzzy numbers in their parametric form, we convert the fractional programming problem in to a single objective linear programming problem in parametric form. We put new fuzzy arithmetic and fuzzy ranking, we obtain the optimal solution the given fully fuzzy linear fractional programming problem without converting to its equivalent crisp linear programming problem. A numerical example is provided to illustrate the efficiency of the proposed method.

In this paper, an efficient approach is presented to solve fully fuzzy linear fractional programming problems. In this approach, first by Alpha-cuts on the objective function and the constraints, the problem is transforming into a bi-objective linear programming problem, and then the resulting problem is converted into two crisp linear programming problems. A numerical example is presented to illustrate the proposed approach. The obtained results demonstrate the acceptable performance of the proposed solution approach.

The fuzzy set theory is being applied massively in many fields these days. One of these is linear programming problems. Kheirfam and Hasani (Sensitivity analysis for fuzzy linear programming problems with fuzzy variables, Advanced Modeling and Optimization, 12 (2010) 257-272), proposed a method for the sensitivity analysis for fuzzy linear programming (FLP) problems with fuzzy variables. There are some important cases that are not considered by Kheirfam and Hasani. In this paper, those cases are considered and numerical examples are solved.

A suggested algorithm to solve triangular fuzzy rough integer linear programming (TFRILP) problems with α-level is introduced in this paper in order to find rough value optimal solutions and decision rough integer variables, where all parameters and decision variables in the constraints and the objective function are triangular fuzzy rough numbers. In real-life situations, the parameters of a linear programming problem model may not be defined precisely, because of the current market globalization and some other uncontrollable factors. In order to solve this problem, a proper methodology is adopted to solve the TFRILP problems by the slice-sum method with the branch-and-bound technique, through which two fuzzy integers linear programming (FILP) problems with triangular fuzzy interval coefficients and variables were constructed. One of these problems is an FILP problem, where all of its coefficients are the upper approximation interval and represent rather satisfactory solutions; the other is an FILP problem, where all of its coefficients are the lower approximation interval and represent completely satisfactory solutions. Moreover, α-level at α = 0.5 is adopted to find some other rough value optimal solutions and decision rough integer variables. Integer programming is used, since a lot of the linear programming problems require that the decision variables be integers. In addition, the motivation behind this study is to enable the decision makers to make the right decision considering the proposed solutions, while dealing with the uncertain and imprecise data. A flowchart is also provided to illustrate the problem-solving steps. Finally, two numerical examples are given to clarify the obtained results.

Chapter two - Linear programming