Please derive the gradient of the loss function in logistic regression (i.e., the logistic loss as shown in below) with respect to w, which is a vector and the model parameter. N L(w) = log(1 + exp (x/w)) yx{w i=1 where (x, y) represents the i-th example, x being data vector for input variables and yi being the label. N is the total number of examples in the data.