In kernel ridge regression (KRR), we convert the data matrix X 2 Rnp to 2 Rnd by running each data xi through the nonlinear transformation function . Correspondingly, the goal in KRR is to minimize nd a weight vector that minimizes the squared loss and the regularization term