SeminarsInformation Visualization for High Dimensional Data:Matrix Visualization
130
reads
Graphical exploration for quantitative/qualitative data acts as the initial yet essential step in modern statistical data analysis. All conventional graphical tools have their own limits: Scatterplot Matrix (SM) is useful for visualizing about only twenty variables; Box-Plot (BP) does not provide interactions between variables; Parallel-Coordinate-Plot (PCP) requires extensive conditioning for extracting overall information. Dimension reduction tools such as Principal Component Analysis (PCA) and MultiDimensional Scaling (MDS) also lose effectiveness when it comes to visual exploration of information structure embedded in very high dimensional data sets. Matrix visualization (MV, Chen (2002); Tien et al. (2008); Wu et al. (2008)) on the other hand can simultaneously explore the associations of up to thousands of subjects, variables, and their interactions, without first reducing dimension. In this lecture I will first briefly introduce the technical background of Matrix visualization (MV, Chen (2002); Tien et al. (2008); Wu et al. (2008)) for continuous, binary, nominal data, and data with cartography links using the Generalized Association Plots (GAP) developed by our laboratory of information visualization. Real applications to scientific problems from biomedical experiments, psychometric studies, and social surveys will then be presented followed by ongoing developments and potential future directions for MV research.
reads
Chun-houh Chen
2009-06-10
08:20:00 - 10:00:00
,
Graphical exploration for quantitative/qualitative data acts as the initial yet essential step in modern statistical data analysis. All conventional graphical tools have their own limits: Scatterplot Matrix (SM) is useful for visualizing about only twenty variables; Box-Plot (BP) does not provide interactions between variables; Parallel-Coordinate-Plot (PCP) requires extensive conditioning for extracting overall information. Dimension reduction tools such as Principal Component Analysis (PCA) and MultiDimensional Scaling (MDS) also lose effectiveness when it comes to visual exploration of information structure embedded in very high dimensional data sets. Matrix visualization (MV, Chen (2002); Tien et al. (2008); Wu et al. (2008)) on the other hand can simultaneously explore the associations of up to thousands of subjects, variables, and their interactions, without first reducing dimension. In this lecture I will first briefly introduce the technical background of Matrix visualization (MV, Chen (2002); Tien et al. (2008); Wu et al. (2008)) for continuous, binary, nominal data, and data with cartography links using the Generalized Association Plots (GAP) developed by our laboratory of information visualization. Real applications to scientific problems from biomedical experiments, psychometric studies, and social surveys will then be presented followed by ongoing developments and potential future directions for MV research.