1. 2.
http://cs229.stanford.edu/notes/cs229-notes3.pdf
Slide credit: J. Sullivan, KTH For more reading refer to R. T. Rockarfeller (1970), Convex Analysis, Princeton University Press.
Slide credit: J. Sullivan, KTH
Slide credit: J. Sullivan, KTH
Slide credit: J. Sullivan, KTH
Slide credit: J. Sullivan, KTH
Slide credit: J. Sullivan, KTH
Slide credit: J. Sullivan, KTH
Slide credit: J. Sullivan, KTH
Slide credit: J. Sullivan, KTH
Slide credit: J. Sullivan, KTH
Slide credit: J. Sullivan, KTH
Slide credit: J. Sullivan, KTH
Slide credit: J. Sullivan, KTH
Slide credit: J. Sullivan, KTH
Slide credit: J. Sullivan, KTH
Slide credit: J. Sullivan, KTH
Slide credit: J. Sullivan, KTH
Slide credit: J. Sullivan, KTH
Slide credit: J. Sullivan, KTH
Slide credit: J. Sullivan, KTH
Slide credit: J. Sullivan, KTH
Slide credit: J. Sullivan, KTH
Slide credit: J. Sullivan, KTH
https://courses.csail.mit.edu/6.867/wiki/images/a/a7/Qp-cvxopt.pdf https://cvxopt.org/examples/tutorial/qp.html
Pros:
a) unique solution
b) automatic selection of support vectors;
Cons:
a) sensitivity to noise
b) kernel selection problem c) constant C?
d) no selection on features
https://en.wikipedia.org/wiki/Representer_theorem
https://en.wikipedia.org/wiki/Representer_theorem
https://en.wikipedia.org/wiki/Representer_theorem