VizStruct Tutorial (2)

 

EM Mixture Plot		EM mixture plot depicts unsupervised data clustering purly based on the visualization distribution of the data set. It offers two visualizations: Cartesian scatter plot: similar to the one offered in basic plot. However, it will also plot the covariance and colors the clustering result. BIC curve and scatter plot matrix: user input a cluster upper bound N>0. A BIC curve and N scatter plot matrix showing how the data set is clustered assuming the number of clusters from 1 to N. Parameters Plot Mode: symbols or numbers in the catesian plot or BIC curve and catesian plot matrix Harmonic: same as in the basic plot Cluster Number: tell VizStruct how much many data clusters in the data set Font Size: same as in the basic plot Plot Color: same as in the basic plot Aspect: same as in the basic plot Preprocess: same as in the basic plot Normalization: same as in the basic plot Sort Dim: same as in the basic plot
Try Yourself		Colored Cartesian plot with cluster number be 2 Click "Default Values" button to reset values Set "Plot Mode" '1. marker' Set "Cluster Number" '2' Set "Font Size" '8' Screen shot of setting Screen shot of plot BIC curve and Cartesian plot matrix upto 8 Click "Default Values" button to reset values Set "Plot Mode" '3. bic' Set "Cluster Number" '2' Set "Font Size" '8' Screen shot of setting Screen shot of plot
Navigation		Tutorial Home Prev TutorialNext Tutorial