As discussed in Support Vector machines chapter, the function that we derive initially for optimisaiton is a non-convex function and hwnce, would get stuck in some local minima on random initialisation.
My doubt is: How do we identify by looking at the function that it is a convex or non-convex function ?
