Hadoop下并行BP神经网络骆马湖水质分类

Based on Parallel BP Neural Network of Classification on Water Quality of Luoma Lake Under Hadoop

研究借助云的计算向数据迁移机制及MapReduce并行处理海量数据的优势,解决BP神经网络在处理大规模样本数据时计算量大、网络训练时间长的瓶颈问题.构建了影响骆马湖水质的多污染因素评价网络模型,在Hadoop下应用并行BP网络算法,实现了对骆马湖水质分类挖掘,挖掘分析结果对骆马湖水质优化及生态修复具有决策支持性意义.

Research the advantage of using the mechanism of computing to data migration and MapReduce parallel processing of massive data,to solve the bottlenecks problem on large amount of computing and network training time when the BP neural network in dealing with a large sample data. Its constructed water quality evaluation model based on the pollution influence factors of Luoma Lake and mined the water quality classification of Luoma Lake by applied the parallel BP algorithm under Hadoop. Mining analysis results is meaningful of decision support for the water quality optimization and ecological remediation of Luoma Lake.