基于遗传算法的数据方体系统设计

决策支持系统实际应用的需求是用户提出的一组频繁查询。为了在系统资源有限的情况下快速有效的回答这些查询,该文提出了基于查询的数据方体系统设计问题。并使用结合了启发式算法的混合遗传算法来优化数据方体系统。在通用数据库上对各种算法进行了实验比较。实验结果表明了该文提出的混合遗传算法在多数情况下可以取得最优解。在不同的系统维护成本的限制下,混合遗传算法都比Harinarayan贪婪算法有效。维护成本越低,效果越明显。

Data cube system design with genetic algorithms

Decision support systems need to respond to frequently asked queries given by users. A data cube system design based on queries was developed to quickly and effectively answer these queries within the system resource limitations. A hybrid genetic algorithm using heuristic information was developed to optimize the data cube system. Tests on a common database showed that the hybrid genetic algorithm quickly reached the optimal solution in most situations. The algorithm is more effective than the greedy algorithm with different system maintenance cost limitations. Since the lower maintenance cost is, the more useful effect is.