在数据仓库中,为选择合适的视图加以实体化,提出一种新的分布估计算法.在解空间随机产生初始群体,根据适应值选择部分好的解集,利用这些优势群体建立概率模型并估计联合概率分布,再从新的概率分布中抽样得到下一代.实验结果表明,该算法能减少查询响应时间和视图维护代价,并且其寻优性能优于经典遗传算法.%To select the appropriate views to be materialized in the data warehouse, a new estimation of distribution algorithm is proposed. It generates initial solutions, the initial population uniformly distributed in the solution space, selects some of the individuals according to their fitness, and builds a posterior probability distribution model of promising solutions New solutions are sampled from the model thus built and fully or in part replaced the old population. Experimental results show that the algorithm reduces the total query response time and maintenance cost, and outperforms canonical genetic algorithm.
展开▼
