Population Sizing for Optimum Sampling with Genetic Algorithms: A Case Study of the Onemax Problem
نویسندگان
چکیده
An experimental and analytical investigation of sampling, which focused on the sizing of population for optimum performance, was performed using the OneMax problem. A fixed-size population and population sizes that vary according to the number of samples were considered in this study. For the case of varying population sizes, population sizing models that ensure proper decision making of the GA and a newly derived model that maximizes the ending generation were studied. A method that provides the optimum sample size is illustrated along with the new population sizing model. The results reveal that maximizing the ending generation of the GA was more successful in maximizing performance.
منابع مشابه
Optimal Sampling and Speed-Up for Genetic Algorithms on the Sampled OneMax Problem
This paper investigates the optimal sampling and the speed-up obtained through sampling for the sampled OneMax problem. Theoretical and experimental analyses are given for three different population-sizing models: the decision-making model, the gambler’s ruin model, and the fixed population-sizing model. The results suggest that, when the desired solution quality is fixed to a high value, the d...
متن کاملGENETIC AND TABU SEARCH ALGORITHMS FOR THE SINGLE MACHINE SCHEDULING PROBLEM WITH SEQUENCE-DEPENDENT SET-UP TIMES AND DETERIORATING JOBS
This paper introduces the effects of job deterioration and sequence dependent set- up time in a single machine scheduling problem. The considered optimization criterion is the minimization of the makespan (Cmax). For this purpose, after formulating the mathematical model, genetic and tabu search algorithms were developed for the problem. Since population diversity is a very important issue in ...
متن کاملA Vibration Damping Optimization Algorithm for Solving the Single-item Capacitated Lot-sizing Problem with Fuzzy Parameters
In this paper, we propose a vibration damping optimization algorithm to solve a fuzzy mathematical model for the single-item capacitated lot-sizing problem. At first, a fuzzy mathematical model for the single-item capacitated lot-sizing problem is presented. The possibility approach is chosen to convert the fuzzy mathematical model to crisp mathematical model. The obtained crisp model is in the...
متن کاملMultiobjective design of sewer networks
The sewer layout in flat areas significantly influences the construction and operational costs as well as reliability of the network performance. To find an optimum design of sewer networks for flat areas, this study presents a multi-objective optimization problem with the objective functions of 1- the cost and 2- the reliability. The reliability criterion is defined as the effect of a clogging...
متن کاملThe Crossover Landscape for the Onemax Problem
In seeking to understand how and why genetic algorithms (GAs) work, attention has been focussed on the landscapes on which they search. While it is relatively simple to analyse the landscapes induced by traditional neighbourhood search operators, the position is considerably complicated for the operators normally used by a GA. Problems which have a single global optimum on a standard`Hamming' l...
متن کامل