改进贝叶斯分类算法的MapReduce并行调度算法  

Parallel scheduling algorithm with improved Bayesian classification algorithm

作　　者：梁宇轩 邢永山 张千[1] 

机构地区：[1]中国石油大学(华东)计算机与通信工程学院,青岛266580

出　　处：《物探化探计算技术》2017年第3期411-415,共5页Computing Techniques For Geophysical and Geochemical Exploration

基　　金：国家自然基金(61671482)

摘　　要：在分析作业划分及现有调度策略的基础上,提出了改进贝叶斯分类算法的作业调度策略,对贝叶斯分类调度算法及MapReduce默认调度方式处理大规模数据时面临的问题进行了阐述,详细地介绍了该改进算法的具体思路和整体流程,描述了该改进算法的具体实现,分析了该调度算法相对其它调度算法的优势。通过实验验证采用改进的贝叶斯调度算法与常用调度算法执行速度进行比较,取得了较好的效果。Based on the analysis of job partitioning and existing scheduling strategies, we proposes a job scheduling strategy that improves the Bayesian classification algorithm, and expounds the problems faced by the Bayesian classifier and the MapRe-duce default scheduling method in dealing with large -scale data in this paper. The concrete idea and the whole process of the improved algorithm, describes the concrete realization of the improved algorithm, and analyzes the advantages of the algorithm compared with other scheduling algorithms are introduced. The experimental results show that the improved Bayesian schedu-ling algorithm is more effective than the conventional scheduling algorithm.

关 键 词：贝叶斯分类算法 MAPREDUCE 地震资料处理 并行调度 

分 类 号：TP319[自动化与计算机技术—计算机软件与理论]