ML Interview Q Series: How are K-Nearest Neighbors and Support Vector Machines different in their fundamental methods, and when might one approach surpass the other?