Search

The Given-Find-Solution (GFS) method is a widely used procedure to systematically understand "what the problem is" in engineering problems. Nevertheless, the GFS method does not provide a clear guide on how to handle uncertainty, feasibility, and preference for design problems. Addressing this issue is crucial since the designer usually requires making assumptions to be able to propose functional solutions. The Given-Find-In Order To-Subjected To (GFIS) method and roadmap are here proposed to systematically state a design problem and requirements using fuzzy desirability functions to handle uncertainty, feasibility, and preference. The GFIS method lists the parameters, variables, specifications, restrictions, and preferences to facilitate the decision-making process. A comparative study was conducted with undergraduate students from Tecnológico de Monterrey enrolled in the course Design Methodologies. The students solved a band brake design problem before and after applying the GFIS method, and their performance was compared. The results showed a significant improvement in their decision-making ability, their capability to find feasible solutions, and the time required to make a design proposal. [ABSTRACT FROM AUTHOR]

This paper proposes a multiclassification algorithm using multilayer perceptron neural network models. It tries to boost two conflicting main objectives of multiclassifiers: a high correct classification rate level and a high classification rate for each class. This last objective is not usually optimized in classification, but is considered here given the need to obtain high precision in each class in real problems. To solve this machine learning problem, we use a Pareto-based multiobjective optimization methodology based on a memetic evolutionary algorithm. We consider a memetic Pareto evolutionary approach based on the NSGA2 evolutionary algorithm (MPENSGA2). Once the Pareto front is built, two strategies or automatic individual selection are used: the best model in accuracy and the best model in sensitivity (extremes in the Pareto front). These methodologies are applied to solve 17 classification benchmark problems obtained from the University of California at Irvine (UCI) repository and one complex real classification problem. The models obtained show high accuracy and a high classification rate for each class. [ABSTRACT FROM AUTHOR]