1. (10 points) Below, L(w) : Red -> R are loss functions. For each L, (i) write down the gradient, and (ii) write down the closed form of the minimizer by setting the gradient to zero. a. L(w) = 1/w - woll3. b. L(w) = ||Aw - bl3, where A is an orthonormal basis, i.e., its columns are unit norm and mutually perpendicular. c. L(w) = 1/Xw - yl/3 + allw/13