TheoremBase

Properties of the Subdifferential on Rn\mathbb{R}^n: Minimisers, Convexity and Closedness, Sums and Positive Multiples

A point minimises f exactly when 0 is a subgradient there; the subdifferential is convex and closed; subgradients of two functions add up to a subgradient of the sum; and subdifferentials scale with positive multiples.

Statement

In the setting of Convex Functions on Rn\mathbb{R}^n: Standing Notation, let C⊆RnC\subseteq\mathbb{R}^{n}, let f,g:C→Rf,g:C\to\mathbb{R} be functions, and let x∈Cx\in C, with subdifferentials ∂f(x)\partial f(x), ∂g(x)\partial g(x) and ∂(f+g)(x)\partial(f+g)(x) of ff, gg and f+g:C→Rf+g:C\to\mathbb{R}, y↦f(y)+g(y)y\mapsto f(y)+g(y).

Minimisers. f(x)≤f(y)f(x)\le f(y) for every y∈Cy\in C if and only if 0∈∂f(x)0\in\partial f(x).

Convexity. ∂f(x)\partial f(x) is a convex subset of Rn\mathbb{R}^{n}.

Closedness. ∂f(x)\partial f(x) is closed.

Sums.

{a+a′: a∈∂f(x), a′∈∂g(x)}⊆∂(f+g)(x).\{a+a':\ a\in\partial f(x),\ a'\in\partial g(x)\}\subseteq\partial(f+g)(x).

Positive multiples. For every real λ>0\lambda>0, the subdifferential of λf:C→R\lambda f:C\to\mathbb{R}, y↦λf(y)y\mapsto\lambda f(y), at xx is

∂(λf)(x)={λa: a∈∂f(x)}.\partial(\lambda f)(x)=\{\lambda a:\ a\in\partial f(x)\}.

Proofs

Log in to submit a proof.

Loading...

Citations

Loading…

Dependencies

Loading…

Related

0 relations

Curated associations between results. These are editable and subjective — they do not replace the dependency graph, which is derived from the references in the text.

No relations recorded yet.

Comments

Log in to comment.

Loading…