Originally posted by: SystemAdmin
>
> What does the "sub" refer to? Why it is not called "gradient" optimization?
Subgradient generalizes the concept of gradient to functions that are not differentiable (but are at least locally convex). For instance, the absolute value function has no gradient at zero, but it has an infinite number of subgradients there (any number between -1 and +1, inclusive).
/Paul
Mathematicians are like Frenchmen: whenever you say something to them, they translate it into their own language, and at once it is something entirely different. (Goethe)
#CPLEXOptimizers#DecisionOptimization