Polarization of an algebraic form

Technique for expressing a homogeneous polynomial in a simpler fashion by adjoining more variables

In mathematics, in particular in algebra, polarization is a technique for expressing a homogeneous polynomial in a simpler fashion by adjoining more variables. Specifically, given a homogeneous polynomial, polarization produces a unique symmetric multilinear form from which the original polynomial can be recovered by evaluating along a certain diagonal.

Although the technique is deceptively simple, it has applications in many areas of abstract mathematics: in particular to algebraic geometry, invariant theory, and representation theory. Polarization and related techniques form the foundations for Weyl's invariant theory.

The technique

The fundamental ideas are as follows. Let f ( u ) {\displaystyle f(\mathbf {u} )} be a polynomial in n {\displaystyle n} variables u = ( u 1 , u 2 , , u n ) . {\displaystyle \mathbf {u} =\left(u_{1},u_{2},\ldots ,u_{n}\right).} Suppose that f {\displaystyle f} is homogeneous of degree d , {\displaystyle d,} which means that

f ( t u ) = t d f ( u )  for all  t . {\displaystyle f(t\mathbf {u} )=t^{d}f(\mathbf {u} )\quad {\text{ for all }}t.}

Let u ( 1 ) , u ( 2 ) , , u ( d ) {\displaystyle \mathbf {u} ^{(1)},\mathbf {u} ^{(2)},\ldots ,\mathbf {u} ^{(d)}} be a collection of indeterminates with u ( i ) = ( u 1 ( i ) , u 2 ( i ) , , u n ( i ) ) , {\displaystyle \mathbf {u} ^{(i)}=\left(u_{1}^{(i)},u_{2}^{(i)},\ldots ,u_{n}^{(i)}\right),} so that there are d n {\displaystyle dn} variables altogether. The polar form of f {\displaystyle f} is a polynomial

F ( u ( 1 ) , u ( 2 ) , , u ( d ) ) {\displaystyle F\left(\mathbf {u} ^{(1)},\mathbf {u} ^{(2)},\ldots ,\mathbf {u} ^{(d)}\right)}
which is linear separately in each u ( i ) {\displaystyle \mathbf {u} ^{(i)}} (that is, F {\displaystyle F} is multilinear), symmetric in the u ( i ) , {\displaystyle \mathbf {u} ^{(i)},} and such that
F ( u , u , , u ) = f ( u ) . {\displaystyle F\left(\mathbf {u} ,\mathbf {u} ,\ldots ,\mathbf {u} \right)=f(\mathbf {u} ).}

The polar form of f {\displaystyle f} is given by the following construction

F ( u ( 1 ) , , u ( d ) ) = 1 d ! λ 1 λ d f ( λ 1 u ( 1 ) + + λ d u ( d ) ) | λ = 0 . {\displaystyle F\left({\mathbf {u} }^{(1)},\dots ,{\mathbf {u} }^{(d)}\right)={\frac {1}{d!}}{\frac {\partial }{\partial \lambda _{1}}}\dots {\frac {\partial }{\partial \lambda _{d}}}f(\lambda _{1}{\mathbf {u} }^{(1)}+\dots +\lambda _{d}{\mathbf {u} }^{(d)})|_{\lambda =0}.}
In other words, F {\displaystyle F} is a constant multiple of the coefficient of λ 1 λ 2 λ d {\displaystyle \lambda _{1}\lambda _{2}\ldots \lambda _{d}} in the expansion of f ( λ 1 u ( 1 ) + + λ d u ( d ) ) . {\displaystyle f\left(\lambda _{1}\mathbf {u} ^{(1)}+\cdots +\lambda _{d}\mathbf {u} ^{(d)}\right).}

Examples

A quadratic example. Suppose that x = ( x , y ) {\displaystyle \mathbf {x} =(x,y)} and f ( x ) {\displaystyle f(\mathbf {x} )} is the quadratic form

f ( x ) = x 2 + 3 x y + 2 y 2 . {\displaystyle f(\mathbf {x} )=x^{2}+3xy+2y^{2}.}
Then the polarization of f {\displaystyle f} is a function in x ( 1 ) = ( x ( 1 ) , y ( 1 ) ) {\displaystyle \mathbf {x} ^{(1)}=\left(x^{(1)},y^{(1)}\right)} and x ( 2 ) = ( x ( 2 ) , y ( 2 ) ) {\displaystyle \mathbf {x} ^{(2)}=\left(x^{(2)},y^{(2)}\right)} given by
F ( x ( 1 ) , x ( 2 ) ) = x ( 1 ) x ( 2 ) + 3 2 x ( 2 ) y ( 1 ) + 3 2 x ( 1 ) y ( 2 ) + 2 y ( 1 ) y ( 2 ) . {\displaystyle F\left(\mathbf {x} ^{(1)},\mathbf {x} ^{(2)}\right)=x^{(1)}x^{(2)}+{\frac {3}{2}}x^{(2)}y^{(1)}+{\frac {3}{2}}x^{(1)}y^{(2)}+2y^{(1)}y^{(2)}.}
More generally, if f {\displaystyle f} is any quadratic form then the polarization of f {\displaystyle f} agrees with the conclusion of the polarization identity.

A cubic example. Let f ( x , y ) = x 3 + 2 x y 2 . {\displaystyle f(x,y)=x^{3}+2xy^{2}.} Then the polarization of f {\displaystyle f} is given by

F ( x ( 1 ) , y ( 1 ) , x ( 2 ) , y ( 2 ) , x ( 3 ) , y ( 3 ) ) = x ( 1 ) x ( 2 ) x ( 3 ) + 2 3 x ( 1 ) y ( 2 ) y ( 3 ) + 2 3 x ( 3 ) y ( 1 ) y ( 2 ) + 2 3 x ( 2 ) y ( 3 ) y ( 1 ) . {\displaystyle F\left(x^{(1)},y^{(1)},x^{(2)},y^{(2)},x^{(3)},y^{(3)}\right)=x^{(1)}x^{(2)}x^{(3)}+{\frac {2}{3}}x^{(1)}y^{(2)}y^{(3)}+{\frac {2}{3}}x^{(3)}y^{(1)}y^{(2)}+{\frac {2}{3}}x^{(2)}y^{(3)}y^{(1)}.}

Mathematical details and consequences

The polarization of a homogeneous polynomial of degree d {\displaystyle d} is valid over any commutative ring in which d ! {\displaystyle d!} is a unit. In particular, it holds over any field of characteristic zero or whose characteristic is strictly greater than d . {\displaystyle d.}

The polarization isomorphism (by degree)

For simplicity, let k {\displaystyle k} be a field of characteristic zero and let A = k [ x ] {\displaystyle A=k[\mathbf {x} ]} be the polynomial ring in n {\displaystyle n} variables over k . {\displaystyle k.} Then A {\displaystyle A} is graded by degree, so that

A = d A d . {\displaystyle A=\bigoplus _{d}A_{d}.}
The polarization of algebraic forms then induces an isomorphism of vector spaces in each degree
A d Sym d k n {\displaystyle A_{d}\cong \operatorname {Sym} ^{d}k^{n}}
where Sym d {\displaystyle \operatorname {Sym} ^{d}} is the d {\displaystyle d} -th symmetric power of the n {\displaystyle n} -dimensional space k n . {\displaystyle k^{n}.}

These isomorphisms can be expressed independently of a basis as follows. If V {\displaystyle V} is a finite-dimensional vector space and A {\displaystyle A} is the ring of k {\displaystyle k} -valued polynomial functions on V {\displaystyle V} graded by homogeneous degree, then polarization yields an isomorphism

A d Sym d V . {\displaystyle A_{d}\cong \operatorname {Sym} ^{d}V^{*}.}

The algebraic isomorphism

Furthermore, the polarization is compatible with the algebraic structure on A {\displaystyle A} , so that

A Sym V {\displaystyle A\cong \operatorname {Sym} ^{\bullet }V^{*}}
where Sym V {\displaystyle \operatorname {Sym} ^{\bullet }V^{*}} is the full symmetric algebra over V . {\displaystyle V^{*}.}

Remarks

  • For fields of positive characteristic p , {\displaystyle p,} the foregoing isomorphisms apply if the graded algebras are truncated at degree p 1. {\displaystyle p-1.}
  • There do exist generalizations when V {\displaystyle V} is an infinite dimensional topological vector space.

See also

  • Homogeneous function – Function with a multiplicative scaling behaviour

References

  • Claudio Procesi (2007) Lie Groups: an approach through invariants and representations, Springer, ISBN 9780387260402 .