Research on parallel liner time selection algorithm on MapReduce model
Wang Yong-gu · Jisuanji gongcheng yu sheji · 2014
The execution efficiency of liner time selection algorithm is relatively low with the increased amount of element.A parallel liner time selection algorithm on mapreduce is proposed.The liner time selection algorithm is redesigned to meet the mapreduce programming model which puts key/value form asits data input.The locally optimal solution is computed in parallel. The locally optimal solution is summaried and the globally optimal solution is gotten.Experiments show that the parallel liner time selection algorithm on mapreduce is with high performance in the face of big data.And the execution efficiency of the algorithm is high with more compute units.