Unsupervised fuzzy tournament selection

Computing and informatics, 2003

2001

In this paper, we propose a dominance-based selection scheme to incorporate constraints into the fitness function of a genetic algorithm used for global optimization. The approach does not require the use of a penalty function and, unlike traditional evolutionary multiobjective optimization techniques, it does not require niching (or any other similar approach) to maintain diversity in the population. The algorithm is validated using two test functions taken from the specialized literature on evolutionary optimization. The results obtained indicate that the approach i s a viable alternative to the traditional penalty function, mainly in engineering optimization problems.

Advanced Engineering Informatics, 2002

In this paper, we propose a dominance-based selection scheme to incorporate constraints into the fitness function of a genetic algorithm used for global optimization. The approach does not require the use of a penalty function and, unlike traditional evolutionary multiobjective optimization techniques, it does not require niching (or any other similar approach) to maintain diversity in the population. We validated the algorithm using several test functions taken from the specialized literature on evolutionary optimization. The results obtained indicate that the approach is a viable alternative to the traditional penalty function, mainly in engineering optimization problems. q (C.A. Coello Coello), emezura@computacion.cs.cinvestav.mx (E. Mezura Montes).

2005

Tournament selection is a popular form of selection which is commonly used with genetic algorithms, genetic programming and evolutionary programming. However, tournament selection introduces a sampling bias into the selection process. We review analytic results and present empirical evidence that shows this bias has a significant impact on search performance. We introduce two new forms of unbiased tournament selection that remove or reduce sampling bias in tournament selection.

Majlesi Journal of Mechatronic Systems, 2017

Tournament selection has been widely used in Genetic Algorithms since it has many advantages such as less time complexity, O(n), applicability for parallel programming, etc. Even if Tournament selection offers more efficiency than other selection methods, it is insufficient to adjust the equilibrium between exploration and exploitation. Because of this insufficiency, it may not convergence to the optimum points closely. The paper introduces a new selection method based on standard Tournament selection. Basically, the new selection method randomly chooses individuals from the population. The method computes fitness values of contestants, then selects the best and the worst individuals between contestants in order to utilize for selection. For some probability value, the selection method decides which individual is chosen, the best or the worst. The new technique overcomes the problem where Tournament is not sufficient by not giving up the exploration. Well-known benchmark functions ...

Journal of Computing in Civil Engineering, 2003

Competition is introduced among the populations of a number of genetic algorithms ͑GAs͒ in solving optimization problems. The aim is to adapt the parameters of the GAs, by altering the resources of the system, so as to achieve better solutions. The evolution of the different populations, having different sets of parameters, is controlled at the level of metapopulation, i.e., the union of populations, on the basis of statistics and trends of the evolution of every population. An overall fitness measure is introduced that incorporates a diversity measure and the required resources to rank the populations. The fuzzy outcome of the conflict among the populations guides the evolution of the different GAs toward better solutions in the statistical sense. The proposed scheme is applied to two different problems-a multimodal function with six global and several near-global optima, and a reliability based optimal design of a simple truss. Numerical results are presented, and the robustness and computational efficiency of the proposed scheme are discussed.

Proceedings of the Genetic and Evolutionary Computation Conference, 2019

Lexicase selection achieves very good solution quality by introducing ordered test cases. However, the computational complexity of lexicase selection can prohibit its use in many applications. In this paper, we introduce Batch Tournament Selection (BTS), a hybrid of tournament and lexicase selection which is approximately one order of magnitude faster than lexicase selection while achieving a competitive quality of solutions. Tests on a number of regression datasets show that BTS compares well with lexicase selection in terms of mean absolute error while having a speed-up of up to 25 times. Surprisingly, BTS and lexicase selection have almost no difference in both diversity and performance. This reveals that batches and ordered test cases are completely different mechanisms which share the same general principle fostering the specialization of individuals. This work introduces an efficient algorithm that sheds light onto the main principles behind the success of lexicase, potentially opening up a new range of possibilities for algorithms to come. CCS CONCEPTS • Computing methodologies → Genetic programming; Selection algorithm;

2017

In our proposed work we have introduces a hybrid algorithm Differential Evolution based on Tournament (TDE) where the best initial population generated through tournament process that has been discuss in details in this paper. The algorithm has been validated with six benchmark classical unconstrained and constrained optimization problems. The experimental results show that our proposed algorithm yield promising results on a test set of unconstrained and constrained optimization problems.

Journal of Physics: Conference Series

Evolutionary algorithm (EA) is a part of evolutionary computing inspired by the theory of biological evolution. Evolutionary algorithms have three parameters that must be defined, namely population size, probability of crossing over, and probability of mutation. The absence of standard rules in setting the value of these parameters becomes a difficulty in utilizing evolutionary algorithms to solve optimization problems so that it can cause early convergence in obtaining local optimum values. This research was conducted to find a way to adjust the value of these parameters by using fuzzy logic to determine the probability of crossing and mutation probability and the method of determining population size based on the best fitness value. This method is called Population Resizing on Fitness Improvement Fuzzy Evolutionary Algorithm (PRoFIFEA). The testing of this method uses the problem of Traveling Salesman Problem (TSP) compared to the standard evolutionary algorithm method which states that the PRoFIFEA method is able to find a more optimal solution.

2002

In this paper, we propose a constraint-handling approach for genetic algorithms which uses a dominance-based selection scheme. The proposed approach does not require the fine tuning of a penalty function and does not require extra mechanisms to maintain diversity in the population. The algorithm is validated using several test functions taken from the specialized literature on evolutionary optimization. The results obtained indicate that the approach can produce reasonable results at low computational costs.