SDSCATTER Interactive scatter plot and visualization of classifier outputs Interactive scatter plot SH=SDSCATTER(DATA) Visualization of classifier outputs SH=SDSCATTER(DATA,P) SH=SDSCATTER(DATA,PD,'roc'); % open scatter and ROC plot SH=SDSCATTER(DATA,PD,'roc',FIG); % open scatter and use ROC plot in FIG INPUT DATA data set P pipeline PD pipeline returning decisions FIG Figure handle of ROC plot opened with SDDRAWROC OUTPUT SH handle of the scatter figure OPTIONS 'callback',FUNC Function handle of a callback (see perclass_example_callback.m) 'callback',FUNC,PAR Pass extra parameter to callback function DESCRIPTION SDSCATTER provides interactive scatter plot and visualization of pipeline outputs. See Scatter menu for interactive functionality. Interactive keyboard commands: cursor keys : change features on x (left/right) and y (up/down) axis =/- : change z-order of classes (for highly overlapping data) > : cycle through classes one at a time l : legend on/off h : hide this class o : show only this class t : this class to top 1:9 : switch to property (see Scatter/Use property menu) a : switch between full data set and subset axes d : show feature distributions f : return to previous sample filter F : reset sample filter (show all samples) SDSCATTER can visualize decisions of classifier P in multi-D spaces or soft outputs in 2D. If PD stores ROC, SDSCATTER with 'roc' option shows also ROC plot and allows inspection of classifier decisions at different operating points. READ MORE http://perclass.com/doc/guide/visualization.html#sdscatter
sdscatter is referenced in examples:
- kb26: Useful tips for confusion matrices
- kb25: Custom callback functions for sdscatter
- kb23: Example on image classification
- kb22: Note on decision tree performance and speed
- kb21: Feature selection in perClass
- kb19: PRTools compatibility
- kb18: How to protect a trained discriminant against outliers?
- kb17: How to optimize three-class classifier in imbalanced problems
- kb16: Visualize the effect of a change of parameters in a trained classifier
- kb15: How to speed up classifiers using the neural network approximation?
- kb12: Detector classifier cascade with ROC analysis
- kb11: Hierarchical classifier: How to build detector-classifier cascade?
- kb10: A step by step construction of a detector
- kb8: How to build a detector in a single line of code?
- kb7: How to convert LIBSVM Support Vector machine into a pipeline?