Canonical: https://knowlpedia.com/convex-analysis/marginal-optimal-value-function/

# Marginal (Optimal Value) Function

The infimum of an objective over a set-valued constraint mapping

Given a [set-valued mapping](https://knowlpedia.com/convex-analysis/set-valued-mapping-multifunction-domain-and-graph-convex-set-valued-mapping/) \(F:X\rightrightarrows Y\) and a function \(\varphi:X\times Y\to\overline{\mathbb{R}}\), the **optimal value (marginal) function** \(\mu:X\to\overline{\mathbb{R}}\) is defined by
\[
\mu(x):=\inf\{\varphi(x,y)\mid y\in F(x)\}, \qquad x\in X.
\]

We use the convention \(\inf(\emptyset):=\infty\), and in the notes it is assumed that \(\mu(x)>-\infty\) for all \(x\in X\).

The marginal function captures "minimize over \(y\) given \(x\)" and is central in parametric optimization and convex analysis; its convexity is addressed in [the convexity theorem for marginal functions](https://knowlpedia.com/convex-analysis/convexity-of-the-marginal-optimal-value-function/).

## Examples

- If \(F(x)\equiv C\) is a fixed nonempty set and \(\varphi(x,y)=g(x,y)\), then \(\mu(x)=\inf_{y\in C}g(x,y)\).
- If \(F(x)\) is the feasible set of a constraint system depending on \(x\), then \(\mu(x)\) is the optimal value of that parameterized problem.
