Abstract | ||
---|---|---|
Correlation analysis is of great significance for exploring the multivariate data sets as it helps researchers toward an in-depth understanding of the complex interactions and relationships among variables. In this paper, we propose a correlation analysis method that identifies salient scalars for multivariate data exploration. We exploit specific mutual information metric to measure the information overlap and analyze the relationships between one scalar and other variables. Moreover, we define the information flow and introduce another metric, influence to quantify the associations among scalars of different variables. Furthermore, we integrate these two information metrics and construct a surprise-influence map for users' interaction to identify the salient scalars. By investigating the relationships among these salient scalars, we analyze the correlations among variables. We demonstrate the applicability and effectiveness of our proposed method by applying it to different data sets. |
Year | DOI | Venue |
---|---|---|
2018 | 10.1109/ACCESS.2018.2864685 | IEEE ACCESS |
Keywords | Field | DocType |
Multivariate data, correlation analysis, specific mutual information, information overlap, information flow | Information flow (information theory),Data mining,Data set,Data visualization,Multivariate statistics,Scalar (mathematics),Computer science,Correlation,Mutual information,Salient,Distributed computing | Journal |
Volume | ISSN | Citations |
6 | 2169-3536 | 0 |
PageRank | References | Authors |
0.34 | 0 | 4 |
Name | Order | Citations | PageRank |
---|---|---|---|
Li Wang | 1 | 0 | 1.69 |
Xiaoan Tang | 2 | 36 | 8.24 |
Junda Zhang | 3 | 0 | 2.03 |
Dongdong Guan | 4 | 4 | 4.79 |