SVM classifier contour plot

I want to calculate the outlier from a dataset using the example 'Detect Outliers using SVM and one-class learning' shown in:

MathWorks

The plot shows several contour plots (with different values) and several support vectors but it is stated that the boundary is where the contour value is zero. Is there any way to plot only the boundary around zero and the respective vectors for just this boundary?

Thank you.

Answer 1

[KSSV]

Read about inpolygon. On knowing the corodinates of contour...you can seperate inside and outside points.

[Angela]

Thank you , i was able to use the 'LevelList' parameter to get the contour I wanted. However it still plots all the support vectors and not just the ones for the particular contour. Any ideas how to achieve that?