大数据环境下并行化先进先出成本算法研究  

Parallel First in First Out Cost Algorithm in Big Data Environment

作　　者：侯宁 HOU Ning(Luthai Textile Co.,Ltd.,Zibo 255000,China)

机构地区：[1]鲁泰纺织股份有限公司信息部

出　　处：《软件导刊》2019年第6期85-88,共4页Software Guide

摘　　要：传统计算机算法在大数据环境下效率较差。为此,从数据处理并行角度出发探索大数据环境下实现先进先出的新算法逻辑,通过先进先出算法实现对成本的有效计算,尤其是提高计算容错性,利用优化的并行化计算模式提高算法时间效率。对传统成本算法与新的并行化先进先出成本算法在实际数据上进行比较实验,结果表明并行化的先进先出成本算法在时间效率上优于传统成本算法,且随着数据量的不断扩大时间效率更加明显,而先进先出的计算模型与传统算法在计算误差上并无扩大,说明并行化的先进先出成本算法在大数据环境下优于传统成本算法。With the popularity of large data environment,the traditional computer algorithms show some problems such as poor effi-ciency in large data environment. This paper explores a new arithmetic logic to implement FIFO in large data environment from the point of data processing parallelism. Through the first-in-first-out(FIFO)algorithm,the cost can be calculated effectively,especially the fault tolerance can be improved in large data mode. By using the optimized parallel computing mode,the time efficiency of the algo-rithm is improved. Experiments on real data show that parallel first-in-first-out cost algorithm is superior to traditional cost algorithm in time efficiency and time efficiency is more obvious with the continuous expansion of data volume,and the first-in-first-out calcula-tion model is also better than the traditional algorithm. There is no expansion in the calculation error. This shows that the parallel first in first out cost algorithm is better than the traditional cost algorithm in big data environment.

关 键 词：大数据环境 先进先出成本算法 并行化计算 时间效率 

分 类 号：TP312[自动化与计算机技术—计算机软件与理论]