Bauer Maximum Principle¶

A convex upper-semicontinuous function on a nonempty compact convex set attains its maximum at an extreme point.

Version: v2 · 2026-09-06 · History
Domain-specific #: 1354
Origin domain: mathematics
Subdomain: convex analysis
Aliases: Bauer's maximum principle, Bauer maximum theorem

Core Idea¶

The Bauer maximum principle states that a convex upper-semicontinuous real-valued function on a nonempty compact convex subset of a locally convex Hausdorff topological vector space attains its maximum at an extreme point. A common elementary statement assumes continuity, which is stronger than upper semicontinuity.^[1]

The recognition invariant is compact convex feasible set + convex objective + upper-semicontinuity + maximum witnessed at an extreme point.

Structural Signature¶

A locally convex Hausdorff topological vector space.
A nonempty compact convex feasible set \(K\).
Extreme points of \(K\).
A real-valued convex objective on \(K\).
Upper semicontinuity, or continuity as a sufficient specialization.
Existence of a global maximum from compactness.
An extreme-point maximizer.
No uniqueness guarantee.
A concave minimum analogue by sign reversal.
Linear objectives satisfying both convex and concave versions.
Dependence on the precise topology and compactness notion.
Connection with Krein–Milman and Choquet theory.

What It Is Not¶

It is not the elementary extreme-value theorem alone; that theorem gives a maximizer but not its location at an extreme point. It is not a minimax theorem and does not exchange optimization order. It does not say every maximizer is extreme or that an extreme-point maximizer is unique.^[2]

Convex functions tend to have informative minima through other machinery; Bauer’s direction is distinctive because it locates a maximum on the extreme boundary.

Scope of Application¶

The principle reduces convex maximization over compact convex sets to extreme points in functional analysis, moment problems, probability measures, differential equations, and mathematical economics. For a polytope it recovers the familiar fact that a linear objective has an optimal vertex, while also covering nonpolyhedral and infinite-dimensional settings under the stated hypotheses.^[3]

Clarity¶

An extreme point cannot be written as a nontrivial convex combination of two distinct points of \(K\). It need not be topologically isolated. Compactness, convexity, objective convexity, and upper semicontinuity perform different jobs; dropping one requires a separate theorem or counterexample.

Manages Complexity¶

The result replaces a search over every feasible point with an existence claim on the extreme boundary. It does not necessarily make computation finite: a compact convex set can have infinitely many extreme points and identifying them may be difficult.

Abstract Reasoning¶

Fix the ambient locally convex topology.
Verify that the feasible set is nonempty, compact, and convex.
Verify objective convexity and upper semicontinuity.
Establish that a maximum is attained.
If a maximizer is not extreme, decompose it into a nontrivial segment.
Use convexity to move to an endpoint without lowering the objective.
Continue with an extreme-point argument justified by compact convex structure.
Report existence separately from uniqueness or computability.

Knowledge Transfer¶

The portable structure is relocating an optimum from a convex bulk to an indecomposable boundary witness. The proposed immediate parent is Optimization.

Examples¶

Polytope. A linear functional on a compact polytope achieves a maximum at some vertex.

Probability measures. On a compact convex family of measures, extreme measures can witness maxima of suitable convex upper-semicontinuous functionals.

Non-example. A continuous concave function can have its maximum strictly inside a convex set.

Structural Tensions¶

Existence versus algorithmic discovery.
Convex bulk versus extreme boundary.
Continuous special case versus upper-semicontinuous theorem.
Extreme-point witness versus nonunique optimum face.
Finite-dimensional intuition versus infinite-dimensional topology.

Structural–Framed Character¶

Boundary localization, indecomposability, and optimum witnessing are structural. Locally convex spaces, compact convex sets, upper semicontinuity, and extreme points are mathematical frame.

Structural Core vs. Domain Accent¶

The portable core is that a convex objective’s maximum can be witnessed by an indecomposable feasible state. The constitutive domain accent is the precise topological-vector-space and convex-analytic hypothesis package.

Optimization is the proposed immediate parent. Convexity, Extremality, Boundary, Compactness, Decomposition, and Existence are related.

The prospective queue contains one strict edge to prime:optimization. No live DAG mutation is authorized.

Relationships to Other Abstractions¶

Current abstraction Bauer Maximum Principle Domain-specific

Parents (1) — more general patterns this builds on

Bauer Maximum Principle is a kind of Optimization Prime

Optimization is the proposed immediate parent.

Hierarchy path (1) — routes to 1 parentless root

Bauer Maximum Principle → Optimization

Neighborhood in Abstraction Space¶

Bauer Maximum Principle sits in a sparse region of the domain-specific corpus (83^rd percentile for distinctiveness): few abstractions share its structure, so a faithful description tends to retrieve it precisely.

Family — Unclustered & Miscellaneous (1565 abstractions)

Nearest neighbors

Computed from structural-signature embeddings · 2026-09-08

Not to Be Confused With¶

Extreme value theorem.
Krein–Milman theorem.
Maximum modulus principle.
Minimax theorem.
A claim that all maxima are extreme.
A finite-vertex enumeration guarantee.
A convex-function minimum principle.

References¶

[1] Robert R. Phelps, Lectures on Choquet’s Theorem, 2^nd ed., Springer, 2001. registry ↩

[2] Charalambos D. Aliprantis and Kim C. Border, Infinite Dimensional Analysis, 3^rd ed., Springer, 2006. registry ↩

[3] Jonathan M. Borwein and Adrian S. Lewis, Convex Analysis and Nonlinear Optimization, 2^nd ed., Springer, 2006. registry ↩

[4] Heinz Bauer, “Minimalstellen von Funktionen und Extremalpunkte,” Archiv der Mathematik 9 (1958): 389–393. registry ↩