Er… for the simple cases I’ve looked at, following the gradient, and solving directly for the maximal-utility point in the space, give the same results, as long as there are no local minima. I should have expected that.
Er… for the simple cases I’ve looked at, following the gradient, and solving directly for the maximal-utility point in the space, give the same results, as long as there are no local minima. I should have expected that.