多维数据可视化汇编.pptVIP

Mathematical ModelingLecture 1 Instructor: Liangfu Lu liangfulv@tju.edu.cn Multi-Dimensional Data Visualization ---Parallel Coordinates * Multi-Dimensional Data Visualizing large information space through relatively small window screen. Visualizing multi-dimensional data (n3) in 2D space. * Axis reconfiguration techniques, such as parallel coordinates and glyphs; Dimensional subsetting, such as scatterplot matrices * parallel coordinates glyphs scatterplot matrices Parallel Coordinates Glyphs Scatterplot Matrices * Background Proposed in 1985 by Alfred Insellberg Good for multi-dimensional data exploration Widely used in information visualization of multi-dimensional data * To represent N dimensional data Set N equidistant vertical axes in parallel Each axis scaled to [min, max] range of attributes Put data to intersects on corresponding axes Connect intersects Example: (0,1,-1,2) * Rationale Cartesian vs. Parallel oordinatesExample 1: Dataset in a Cartesian coordinate Same dataset in parallel coordinates * Example 2: Air traffic control Cartesian Coordinates Parallel Coordinates * * Good Example Parallel Coordinates: Advantages Multi-dimensional data can be visualized in two dimensional space with low complexity. Each variable is treated uniformly. Relations within multi-dimensional data can be discovered (“data mining”). * * Bad Examples Parallel Coordinates: Disadvantages Axes are too close as dimensions increase. Clutter can reduce perceived information. Varying the order of axes scale, although indicating different patterns, may cause confusion. * Parallel Coordinates Visualization Summary Each data point is mapped to a line (Duality) Similar points correspond to similar lines(Correlation) Effective exploration and clustering(Pattern) Problems: order of axes, limit to 1~20 dimensions *