The weights w; are estimated by minimizing a regularized mean square error: min w w))) (y = f(x;; w)) + wKw, i=1 where w is the column vector formed by w = [w] and K is the kernel matrix. Please derive the optimal solution of w using matrix inverseion