Bernstein Polynomial Basis Overview

• Bernstein polynomial basis is a family of nonnegative, partition-of-unity polynomials defined on compact intervals with strong approximation and geometric properties.
• They facilitate shape-preserving representations in Bézier curves, spectral solutions for PDEs, and global optimization with convergent relaxations.
• Their combinatorial and algebraic structure enables efficient change-of-basis operations and robust numerical conditioning in practical computational applications.

The Bernstein polynomial basis is a family of nonnegative, partition-of-unity polynomial functions parameterized by degree, defined on compact intervals or multidimensional boxes. It is characterized by its combinatorial binomial structure, strong approximation properties, geometric and probabilistic interpretations, and its central role in applications ranging from geometric design to numerical analysis, global optimization, and the theory of polynomial identities. The defining univariate form on $[a, b]$ is

$$B_{i,n}(x) = \binom{n}{i} \frac{(x-a)^i (b-x)^{n-i}}{(b-a)^n} \,,\qquad i=0,\dots,n,\,\, x\in[a,b]$$

which on $[0,1]$ simplifies to $B_{i,n}(x) = \binom{n}{i} x^i (1-x)^{n-i}$. For the multivariate unit cube $[0,1]^d$, the basis generalizes by tensor product. Bernstein bases are fundamental in computer-aided geometric design (Bézier curves/surfaces), spectral and collocation methods for PDEs, polynomial optimization, and serve as a canonical link between combinatorial, analytical, and algebraic viewpoints.

1. Definition, Structure, and Core Properties

The $n$-th degree univariate Bernstein basis on $[a,b]$ consists of functions

$$B_{i,n}(x) = \binom{n}{i} \frac{(x-a)^i (b-x)^{n-i}}{(b-a)^n}$$

For $[0,1]$ this is $B_{i,n}(x) = \binom{n}{i} x^i (1-x)^{n-i}$. Core properties include:

• Positivity: $$B_{i,n}(x) = \binom{n}{i} \frac{(x-a)^i (b-x)^{n-i}}{(b-a)^n} \,,\qquad i=0,\dots,n,\,\, x\in[a,b]$$0 for all $$B_{i,n}(x) = \binom{n}{i} \frac{(x-a)^i (b-x)^{n-i}}{(b-a)^n} \,,\qquad i=0,\dots,n,\,\, x\in[a,b]$$1 and $$B_{i,n}(x) = \binom{n}{i} \frac{(x-a)^i (b-x)^{n-i}}{(b-a)^n} \,,\qquad i=0,\dots,n,\,\, x\in[a,b]$$2.
• Partition of Unity: $$B_{i,n}(x) = \binom{n}{i} \frac{(x-a)^i (b-x)^{n-i}}{(b-a)^n} \,,\qquad i=0,\dots,n,\,\, x\in[a,b]$$3 for all $$B_{i,n}(x) = \binom{n}{i} \frac{(x-a)^i (b-x)^{n-i}}{(b-a)^n} \,,\qquad i=0,\dots,n,\,\, x\in[a,b]$$4.
• Endpoint Interpolation:

$$B_{i,n}(x) = \binom{n}{i} \frac{(x-a)^i (b-x)^{n-i}}{(b-a)^n} \,,\qquad i=0,\dots,n,\,\, x\in[a,b]$$5

• Recurrence:

$$B_{i,n}(x) = \binom{n}{i} \frac{(x-a)^i (b-x)^{n-i}}{(b-a)^n} \,,\qquad i=0,\dots,n,\,\, x\in[a,b]$$6

with $$B_{i,n}(x) = \binom{n}{i} \frac{(x-a)^i (b-x)^{n-i}}{(b-a)^n} \,,\qquad i=0,\dots,n,\,\, x\in[a,b]$$7 for $$B_{i,n}(x) = \binom{n}{i} \frac{(x-a)^i (b-x)^{n-i}}{(b-a)^n} \,,\qquad i=0,\dots,n,\,\, x\in[a,b]$$8 or $$B_{i,n}(x) = \binom{n}{i} \frac{(x-a)^i (b-x)^{n-i}}{(b-a)^n} \,,\qquad i=0,\dots,n,\,\, x\in[a,b]$$9.

• Support and Degree: Each $[0,1]$0 is a polynomial of degree $[0,1]$1, vanishing outside $[0,1]$2.
• Linear Independence: The set $[0,1]$3 forms a basis for $[0,1]$4, the polynomials of degree at most $[0,1]$5.

On $[0,1]$6, explicit examples are: $[0,1]$7 The basis readily generalizes to multidimensional domains via tensor products and barycentric coordinates for simplices and tetrahedra (Mirkov et al., 2012, Ainsworth et al., 2018).

2. Analytic Identities: Generating Functions, Recursion, and Differential Relations

A unified generating function for the Bernstein basis is

$[0,1]$8

Specializing to $[0,1]$9, $B_{i,n}(x) = \binom{n}{i} x^i (1-x)^{n-i}$0 yields the classical basis. This approach enables:

• Recursion:

$B_{i,n}(x) = \binom{n}{i} x^i (1-x)^{n-i}$1

• Differentiation:

$B_{i,n}(x) = \binom{n}{i} x^i (1-x)^{n-i}$2

• Monomial Expansion:

$B_{i,n}(x) = \binom{n}{i} x^i (1-x)^{n-i}$3

• Alternating Sum:

$B_{i,n}(x) = \binom{n}{i} x^i (1-x)^{n-i}$4

Additionally, explicitly constructed degree-elevation, subdivision, and convolution formulas arise naturally (Simsek, 2010, Simsek, 2011).

The generating function formalism further yields expressions for moments and connects the basis to binomial and Poisson distributions. For $B_{i,n}(x) = \binom{n}{i} x^i (1-x)^{n-i}$5 interpreted as the probability of $B_{i,n}(x) = \binom{n}{i} x^i (1-x)^{n-i}$6 "successes" in $B_{i,n}(x) = \binom{n}{i} x^i (1-x)^{n-i}$7 Bernoulli trials with success probability $B_{i,n}(x) = \binom{n}{i} x^i (1-x)^{n-i}$8, one obtains mean and variance $B_{i,n}(x) = \binom{n}{i} x^i (1-x)^{n-i}$9, $[0,1]^d$0 (Simsek, 2010).

3. Algebraic Structure and Change-of-Basis Theory

Any polynomial $[0,1]^d$1 of degree $[0,1]^d$2 can be written uniquely in the Bernstein basis: $[0,1]^d$3 The change of basis between monomials and Bernstein polynomials is upper- (descending) or lower- (ascending) triangular, with entries given by explicit binomial or generalized hypergeometric sums (Wolfram, 2022): $[0,1]^d$4 where

$[0,1]^d$5

These matrices admit fast computation via recurrence, hypergeometric evaluation, or Lagrange-polynomial interpolation (Wolfram, 2022). Truncation, alternation, and superposition functorially generate new bases and groupoid structures linking Bernstein and other classical families (Zernike, Chebyshev, Legendre).

4. Numerical and Approximation Properties

Key properties underpinning applications include:

• Nonnegativity and Partition of Unity: The representation $[0,1]^d$6 ensures $[0,1]^d$7 if all $[0,1]^d$8. This convex hull property is foundational in geometric modeling and strict global optimization (Amorese et al., 10 Jun 2025, Sassi et al., 2015).
• Spectral/Exponential Convergence: Bernstein bases exhibit exponential decay of $[0,1]^d$9 error in spectral/collocation solvers for PDEs (e.g., Poisson, Helmholtz, biharmonic equations) up to a degree limit where floating-point error dominates (Mirkov et al., 2012). Example: On $n$0, for $n$1, error decays from $n$2 ($n$3) to $n$4 ($n$5).
• Variation-diminishing and Shape-preservation: Linear combinations preserve nonnegativity and avoid spurious oscillations (key in spline and geometric applications).
• Numerical Conditioning: The basis is well-conditioned for moderate $n$6 in both integral-equation solvers and LP relaxations of polynomial-optimization problems (Shirin et al., 2013, Amorese et al., 10 Jun 2025, Sassi et al., 2015). No adverse ill-conditioning is reported for $n$7 up to 20–30 in practical PDE/ODE solvers or optimization relaxations.

5. Generalizations: Multivariate, Modal, and Orthonormal Bernstein Bases

• Multivariate Bernstein Basis: For $n$8 and degree $n$9,

$[a,b]$0

The same nonnegativity, partition, and convex hull properties extend, forming the backbone of global optimizers, finite element methods, and box-constrained certifications (Sassi et al., 2015, Ainsworth et al., 2018).

• Dual and Modal Bases: Dual Bernstein polynomials $[a,b]$1 provide $[a,b]$2 biorthogonality to the standard basis, and their linear combinations construct modal test functions for Petrov-Galerkin methods, enforcing boundary/regularity constraints while yielding banded, sparse linear systems (Jani et al., 2016).
• Orthonormal Bernstein Basis: There exist explicit lower-triangular transforms $[a,b]$3 yielding an orthonormal set $[a,b]$4 in $[a,b]$5:

$[a,b]$6

These are eigenfunctions of a Sturm-Liouville operator with explicit spectrum and form the core of generalized Fourier-Bernstein expansions (Bellucci, 2014).

6. Nonnegativity Certification, Optimization, and Computational Methods

Classical nonnegativity certificate: If all Bernstein coefficients are nonnegative, the polynomial is nonnegative on $[a,b]$7. Newer, less restrictive certificates replace each coefficient's lower bound using geometric means of its neighbors (generalizing to matrix-valued cases via matrix geometric means). This technique provides dramatically fewer subdivisions and works efficiently in scalar and matrix scenarios compared to full SOS relaxations (Harris et al., 2023).

For polynomial optimization over $[a,b]$8:

• The Bernstein convex hull property ensures $[a,b]$9 is bounded between $$B_{i,n}(x) = \binom{n}{i} \frac{(x-a)^i (b-x)^{n-i}}{(b-a)^n}$$0 and $$B_{i,n}(x) = \binom{n}{i} \frac{(x-a)^i (b-x)^{n-i}}{(b-a)^n}$$1.
• Linear programming relaxations can be defined using these bounds, further tightened by incorporating degree reduction and upper-bound constraints, yielding provably convergent hierarchies (Bernstein inequalities) (Sassi et al., 2015, Amorese et al., 10 Jun 2025).
• The Bernstein-based LP approach contrasts with the Sum-of-Squares/SDP approach: the Bernstein relaxations provide asymptotically complete relaxations and outpace SOS in convergence rate but may be less computationally feasible in higher dimensions due to explosion in the number of constraints (Amorese et al., 10 Jun 2025).

7. Applications: Integral Equations, PDEs, Bézier Geometry, and Beyond

• Galerkin and Collocation Methods: In Fredholm and boundary value problems, Bernstein bases provide trial spaces leading to well-conditioned, rapidly convergent schemes. In the Galerkin method, the system matrix consists of integrals over products of Bernstein polynomials and problem data (Shirin et al., 2013, Mirkov et al., 2012, Jani et al., 2016).
• Computer-Aided Geometric Design (CAGD): The geometric control offered by the Bernstein basis underpins Bézier curves and surfaces, with direct manipulation of control-point polygon/mesh guaranteeing shape-preserving behavior and endpoint interpolation (AlQudah, 2015, Bellucci, 2014).
• Finite Elements on Tetrahedra: Multivariate Bernstein bases in barycentric form enable efficient and exactly computable assembly of mass and stiffness matrices for $$B_{i,n}(x) = \binom{n}{i} \frac{(x-a)^i (b-x)^{n-i}}{(b-a)^n}$$2, $$B_{i,n}(x) = \binom{n}{i} \frac{(x-a)^i (b-x)^{n-i}}{(b-a)^n}$$3, $$B_{i,n}(x) = \binom{n}{i} \frac{(x-a)^i (b-x)^{n-i}}{(b-a)^n}$$4, and $$B_{i,n}(x) = \binom{n}{i} \frac{(x-a)^i (b-x)^{n-i}}{(b-a)^n}$$5 elements, exploiting combinatorial identities for fast product and integration (Ainsworth et al., 2018).
• Spectral and Modal Methods: Modal and dual Bernstein bases enable sparse and highly accurate solvers for fractional and integer PDEs, especially when boundary and regularity conditions require specific enforcement (Jani et al., 2016, Mirkov et al., 2012).
• Global Optimization and Safety Verification: Bernstein relaxations are central for bounding the minimum or certifying safety of high-degree polynomials over boxes, with certified convergence rates and connections to probabilistic safety-barrier verification (Sassi et al., 2015, Amorese et al., 10 Jun 2025).

In all these domains, the intrinsic positivity, partitioning, and algebraic structure of the Bernstein polynomial basis deliver a robust, geometric, and computationally tractable framework that bridges combinatorics, analysis, algebra, and computational science (Shirin et al., 2013, Mirkov et al., 2012, Simsek, 2010, Sassi et al., 2015, Amorese et al., 10 Jun 2025).