Population Incomelevel Of The Forecast Based On HBase

With the arrival of the era of big data, various industries and fields began to use big data processing technology, the mass data in the information and knowledge mining, to provide a basis for human social activities. Big data technology not only can store huge amounts of data sets, but also provides distributed parallel architecture processing technology, which can improve the data throughput. In the face of massive population income data sets, the use of large data processing technology will improve the processing speed and efficiency of data.This article after the analysis of the current mainstream data processing technology, the Hadoop as big data processing platform, HBase as large data sets stored in the database, through the use of the MapReduce programming model to achieve data on the income of the population, mainly realizes the following work:Analysis of open source Hadoop platform technology, focus on the HBase column storage database, including data model, design of the table, table operation and data import, and through the installation of Hadoop and HBase to realize data storage of the data on the income of the population.The application of naive Bayes algorithm in population income data is studied. For storage in HBase census income data set, through the realization of MapReduce for naive Bayes algorithm to realize the data processing, including data discretization, the algorithm model of training, the algorithm model of the evaluation process implementation of MapReduce.Through the naive Bayesian algorithm of MapReduce three processes, for storage in HBase census income data for processing and analysis, the experimental results are given, for the future in terms of dealing with massive population and income data provides a good reference value.