Take the 2-minute tour ×
MathOverflow is a question and answer site for professional mathematicians. It's 100% free, no registration required.

Let $E\subset\mathbb{R}^n$ be a convex set, $u:E\to\mathbb{R}$ a convex function and $B\subset E$ a Borel set.

We define the (multivalued) gradient \begin{array}{rccl} \nabla [u]:& \mathbb{R}^n & \to & \mathcal{P}((\mathbb{R}^n)^\vee)\\ & {a} & \mapsto & \{ \xi\in(\mathbb{R}^n)^\vee \mid {x}\mapsto \langle \xi, {x}\rangle -u({x}) \text{ attains its absolute maximum at }{a}\}. \end{array}

Informally, $\nabla [u]({a})$ is the set of slopes of all hyperplanes touching the graph at the point $({a},u({a}))\in \mathbb{R}^n\times\mathbb{R}$ of the function and staying below.

Also we define $\nabla[u](B):=\bigcup_{{x}\in B} \nabla[u]({x})$.

Finally the Monge-Ampère measure of $u$ of $B$ is $M[u](B):=\lambda_L(\nabla[u](B))$, where $\lambda_L$ is the Lebesgue measure. In some sense it measures the total convexity of $u$ in $B$.

Is it easy from this definition to see that $M:${convex functions}$\to${$\mathbb{R}^n$positive measures} is continuous for the locally uniform convergence topology and the weak topology, respectively?

(I am reading the article of Passare and Rullgard "Amoebas, Monge-Ampère measures, and triangulations of the Newton polytope")

share|improve this question
add comment

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Browse other questions tagged or ask your own question.