Genetic Algorithms in Search, Optimization and Machine Learning
Genetic Algorithms in Search, Optimization and Machine Learning
MINITAB Guide to Statistics
An elitist genetic algorithm for multiobjective optimization
Metaheuristics
| Hi-index | 0.00 |
A major disadvantage of using a genetic algorithm for solving a complex problem is that it requires a relatively large amount of computational time to search for the solution space before the solution is finally attained. Thus, it is necessary to identify the tradeoff between the algorithm stopping criteria and the algorithm performance. As an effort of determining the tradeoff, this paper examines the relationship between the algorithm performance and algorithm stopping criteria. Two algorithm stopping criteria, such as the different numbers of unique schedules and the number of generations, are used, while existing studies employ the number of generations as a sole stopping condition. Elitist genetic algorithm is used to solve 30 projects having 30-Activity with four renewable resources for statistical analysis. The relationships are presented by comparing means for algorithm performance measures, which include the fitness values, total algorithm runtime in millisecond, and the flatline starting generation number.