Canonical: https://knowlpedia.com/convex-analysis/biconjugate/

# Biconjugate

The Fenchel conjugate of a function's conjugate, central to lower-semicontinuous convex relaxation.

A **biconjugate** of an extended-real-valued [function](https://knowlpedia.com/shared-foundations/function/) \(f:\mathbb{R}^n\to(-\infty,+\infty]\) is the function
\[
f^{**}=(f^*)^*,
\]
where \(f^*\) is the [Fenchel conjugate](https://knowlpedia.com/convex-analysis/convex-conjugate-fenchel/) of \(f\).

If \(f\) is proper and has an affine minorant, then \(f^{**}\) is the greatest lower-semicontinuous convex function bounded above by \(f\). In particular, \(f^{**}\le f\) pointwise. The [Fenchel–Moreau theorem](https://knowlpedia.com/convex-analysis/fenchel-moreau-theorem/) states that a proper function satisfies \(f=f^{**}\) exactly when it is lower-semicontinuous and convex.

Here “lower-semicontinuous and convex” is also commonly called [closed convex](https://knowlpedia.com/convex-analysis/closed-convex-function/).

## Examples

- If \(f\) is proper, lower-semicontinuous, and convex—for instance, a norm—then \(f^{**}=f\).
- If \(C\subseteq\mathbb{R}^n\) is nonempty and \(\delta_C\) is its indicator function, then \(\delta_C^{**}=\delta_{\overline{\operatorname{conv}}(C)}\). For \(C=\{-1,1\}\subset\mathbb{R}\), this gives \(\delta_C^{**}=\delta_{[-1,1]}\).
