基于单调向量空间的大数据分析方法研究

Research on Big Data Analysis Method Based on Monotonic Vector Space

在获取海量数据后,大数据分析方法通过寻找数据之间的关系来解决问题,数据间的关系虽然很难被抽象出来,但却客观存在,并可以表示为某些函数。本文通过引入单调向量空间,分析海量数据可能有用的因素,构成多维单调向量空间模型,即通过单调映射函数来表示大数据关系。然后,对该模型进行算法分析,通过分割算法能够得到满足需求的数据集合,通过筛选算法和敏感区域算法,能够得到关键因素和关键区域,通过可视化方法能够降维显示大数据。本文的方法可以为大数据的处理与分析提供借鉴。

For big data’s research,we are looking for the relationship among data in order to solve problems,because everyone believes that there always exists an objective function about arguments and the dependent variable.After learning about Monotonic Vector Space(MVS),this paper found that MVS supports monotonic mappings which represent the function above.It is assumed that the potentially useful factors constitute multiple dimensions.Then the algorithm of MVS can be used for data analysis.For example,dividing algorithm was used to seek the intersection of multiple,and sequential bifurcation algorithm and the algorithm of sensitive areas were also used to screen important index or important areas,and the big data could be reduced in dimension through the visualization method.The method of this paper can provide reference for the processing and analysis of big data.