Bernstein polynomial¶

A polynomial represented in the Bernstein basis, whose nonnegative partition-of-unity weights support stable approximation, shape preservation, and Bézier geometry.

Version: v2 · 2026-09-19 · History
Domain-specific #: 3439
Domain group: Formal Sciences
Origin domain: Mathematics
Subdomain: Approximation Theory and Computer Graphics → Mathematics

Core Idea¶

Degree-n Bernstein basis functions combine binomial coefficients with xᵏ(1−x)ⁿ⁻ᵏ; weighted control values define a polynomial on a normalized interval. The basis weights are nonnegative and sum to one, making each value a convex combination of coefficients; degree elevation and de Casteljau recursion preserve the curve while improving representation and evaluation. The abstraction is therefore identified by a declared carrier, a transformation or constraint over that carrier, and an invariant that tells an analyst whether the named structure is genuinely present.

How would you explain it like I'm…

Fair-Share Mixing Curve

Imagine a few friends each holding a number, and a slider that moves from one end to the other. At each spot on the slider, every friend gets a share of a vote, the shares always add up to one whole, and the answer is the mixed-together number. Near one end the first friend matters most, near the other end the last friend does. That smooth mixing recipe is a Bernstein polynomial.

Blending Numbers into a Curve

A Bernstein polynomial is a way to make a smooth curve from a short list of numbers called control values. You move along a stretch from 0 to 1, and at every point each control value gets a weight. The weights are never negative and always add up to exactly 1, so the curve is always a fair blend of the control values and never shoots past the biggest or below the smallest. At the start, the first control value gets all the weight; at the end, the last one does.

Convex-Blend Polynomial Basis

A Bernstein polynomial of degree n is built on the interval from 0 to 1 using n+1 basis functions, each a binomial coefficient times x to a power times (1−x) to the remaining power. You multiply each basis function by a control value (a coefficient) and add them up. Because the basis functions are all nonnegative and add to one at every x, each value of the polynomial is a weighted average of the control values, so the curve stays inside their range. You can raise the degree (add more control values) without changing the curve, and you can evaluate it by repeatedly taking in-between points, a method called de Casteljau's algorithm. It is important to keep straight which thing you mean: the basis functions, a polynomial written in this basis, or the 'Bernstein approximation' of some other function.

The degree-n Bernstein basis on [0,1] consists of b_{k,n}(x) = C(n,k) x^k (1−x)^(n−k) for k = 0,…,n, and a Bernstein polynomial is a linear combination Σ β_k b_{k,n}(x) with coefficients (control values) β_k. The basis functions are nonnegative on the interval and form a partition of unity, so each value of the polynomial is a convex combination of the coefficients; this gives bounds by the coefficient range and underpins the geometric intuition in Bézier curves. Degree elevation rewrites the same polynomial in the degree-(n+1) basis with new coefficients, and de Casteljau's algorithm evaluates it by repeated linear interpolation between adjacent coefficients, both leaving the curve unchanged. Using the idea correctly requires stating the interval, the degree, the normalization of the basis, and the coefficients. It also requires distinguishing three things that share the name: the basis functions, a polynomial represented in that basis, and the Bernstein approximation operator that builds such a polynomial from samples of a function.

Scope of Application¶

Bernstein polynomial belongs to approximation theory and computer graphics and is useful where the analyst can specify the typed approximation theory and computer graphics carrier, defining objects and relations, parameters, conventions, evidence, boundary cases, and comparison targets, then evaluate the interval, degree, Bernstein basis normalization, coefficient or control values, and distinction between basis functions, representation, and Bernstein approximation operator are explicit. The scope is broad within that domain but bounded by the need for the interval, degree, Bernstein basis normalization, coefficient or control values, and distinction between basis functions, representation, and Bernstein approximation operator are explicit. The entry records a descriptive analytical identity; practical use requires the governing domain's evidence, standards, and safety obligations.

Clarity¶

The abstraction clarifies a crowded vocabulary by making the interval, degree, Bernstein basis normalization, coefficient or control values, and distinction between basis functions, representation, and Bernstein approximation operator are explicit the center of the account. A claim should name the carrier, the governing operation or relation, the applicable assumptions, and the recognition test. A bare label is insufficient because the name Bernstein polynomial can be used for a formal identity, an implementation, or a neighboring result unless carrier and convention are stated.

Manages Complexity¶

Without the abstraction, an analyst must reason directly over many local details: the carrier roles, admissibility assumptions, competing conventions, derived invariants, boundary cases, and proof or validation obligations specific to Bernstein polynomial. Bernstein polynomial compresses them into the roles in the structural signature. That compression permits comparison across instances without erasing the variables that determine validity. It also exposes which details may be varied safely and which are constitutive.

Abstract Reasoning¶

Identify the carrier. State what the elements, states, objects, or observations are: the typed approximation theory and computer graphics carrier, defining objects and relations, parameters, conventions, evidence, boundary cases, and comparison targets. Reject examples whose alleged carrier belongs to a different problem.
Lock the constitutive rule. Express the interval, degree, Bernstein basis normalization, coefficient or control values, and distinction between basis functions, representation, and Bernstein approximation operator are explicit independently of one notation or implementation.

Knowledge Transfer¶

Knowledge transfers strongly among subfields of approximation theory and computer graphics because they reuse the typed approximation theory and computer graphics carrier, defining objects and relations, parameters, conventions, evidence, boundary cases, and comparison targets, The basis weights are nonnegative and sum to one, making each value a convex combination of coefficients; degree elevation and de Casteljau recursion preserve the curve while improving representation and evaluation., and type the carrier, state every parameter and convention in the definition, test that the interval, degree, Bernstein basis normalization, coefficient or control values, and distinction between basis functions, representation, and Bernstein approximation operator are explicit, compare the nearest accepted identity, and report counterexamples, uncertainty, and limiting cases.

Relationships to Other Abstractions¶

Current abstraction Bernstein polynomial Domain-specific

Parents (1) — outgoing typed relations

Bernstein polynomial is a kind of Representation Prime

It remains its own entry because its identity is fixed by the interval, degree, Bernstein basis normalization, coefficient or control values, and distinction between basis functions, representation, and Bernstein approximation operator.

Hierarchy path (1) — routes to 1 parentless root

Bernstein polynomial → is a kind of Representation → presupposes Abstraction

Neighborhood in Abstraction Space¶

Bernstein polynomial sits in a crowded region of the domain-specific corpus (24^th percentile for distinctiveness): several abstractions share nearly its structure, so a description that fits it tends to fit its neighbors too.

Family — Unclustered & Miscellaneous (2769 abstractions)

Nearest neighbors

Computed from structural-signature embeddings · 2026-10-25