Transactions on Computational Collective Intelligence XVII
Transactions on Computational Collective Intelligence XVII
复制标题
计算集体智能汇刊 XVII
DOI：
10.1007/978-3-662-44994-3_12
复制
发表时间：
2014
期刊：
影响因子：
--
通讯作者：
Sapin E
中科院分区：
文献类型：
--
作者：
Sapin E
对大数据的分析，特别是来自生物科学的大数据的分析，对用于分析这些数据的方法提出了独特的挑战。全基因组关联研究中使用的数据集可能具有非常多的变量/维度（例如400，000+），因此需要对标准方法进行修改以使其正确运行。各种方法可用于此类问题，其中蚁群优化是一种有前途的方法，其灵感来自于蚂蚁在自然界中找到最短路径的方式。路径的选择传统上使用轮盘赌轮，它适用于较小维度的问题，但当考虑更多变量时就会崩溃。在本文中，提出了一种基于子集的比赛选择ACO方法，表现出优于轮盘赌轮为基础的方法，运筹学的高维问题的性能的最终解决方案和执行时间的问题，从文献。
The analysis of big data, particularly from the biosciences, provides unique challenges to the methods used to analyse such data. Datasets such as those used in genome-wide association studies can have a very high number of variables/dimensions (e.g. 400,000+) and therefore modifications are required to standard methods to allow them to function correctly.A variety of methods can be used for such problems, among them ant colony optimisation is a promising method, inspired by the way in which ants find the shortest path in nature. The selection of paths traditionally uses a roulette wheel which works well for problems of smaller dimensionality but breaks down when higher numbers of variables are considered. In this paper, a subset-based tournament selection ACO approach is proposed that is shown to outperform the roulette wheel-based approach for operations research problems of higher dimensionality in terms of the performance of the final solutions and execution time on problems taken from the literature.