Problem C [9 points]: Leaving squared loss behind, let's focus on log loss and hinge loss. Consider the set of points S={(21,3),(2,2),(3,1)} in 2D space, shown below, with labels (1,1,1) respectively. Given a linear model with weights w0=0,w1=1,w2=0 (where w0 corresponds to the bias term), derive the gradients wLhinge and wLlog of the hinge loss and log loss, and calculate their values for each point in S. - Hinge loss: Lhinge=max(0,1ywTx) - logloss:Llog=ln(1+eywTx)