基于Hadoop物联网数据挖掘的算法分析与应用

Analysis and application of data mining algorithm for Internet of Things based on Hadoop

介绍了物联网数据处理的若干关键技术,如大数据采集、大数据存储、大数据的分析与挖掘等。以Hadoop为平台对物联网数据进行挖掘与分析,为了提高处理庞大数据的实效性,基于Map Reduce架构采用了朴素贝叶斯分类算法、K-modes聚类算法以及ECLAT算法。分析认为,应用这三类算法,提高了数据分类效率,优化了类内对象之间的相似性以及类间对象之间的关联性,为更高效的数据挖掘提供了很好的思路。

Some key technologies of data processing for Internet of Things are introduced, such as big data acquisition, big data storage, big data analysis and mining. In this paper, the data of Internet of Things is excavated and analyzed on Hadoop platform,In order to improve the effectiveness of large data processing, Naive Bayesian classification algorithm, K-modes clustering algorithm and ECLAT algorithm are adopted in Map Reduce framework. The analysis shows that the application of these three kinds of algorithms improves the efficiency of data classification, optimizes the similarity among the objects in the class and the correlation among the objects between classes, and provides a good idea for more efficient data mining.