Is there a general classification of when linear monoids (i.e. monoids given by the solutions to a linear integral equation) are finitely generated in $\mathbb{N}^k$ or in $\mathbb{Z}^k$?

Algebraic geometry background:

The claim that quasi-projective $A$-schemes are finite type can be reframed as a statement about finite generation of certain monoids in $\mathbb{N}^k$, in the following way. We would like to show that if $S_\bullet$ is a finitely generated graded $A$-algebra, say with generators $(s_1,\ldots, s_k)$, then for any $f$ homogeneous of positive degree, $$ ((S_\bullet)_f)_0 $$ is finitely generated as an $A$-algebra.

This subring is generated over $A$ by all monomials of the form $$ \left\{\frac{s_{1}^{e_1} \cdots s_{k}^{e_k}}{f^n} : n \deg f = \sum_{j = 1}^k e_j \deg s_j\right\}. $$ The tricky algebraic fact to prove is that a finite subset of these monomials suffices to generate the rest under multiplication. Equivalently, we would like to know that the additive monoid $$ \left\{(e_1, \ldots, e_k, n) \in \mathbb{N}^{k+1} : n \deg f = \sum_{j = 1}^k e_j \deg s_j \right\} $$ is finitely generated for arbitrary positive coefficients $\deg s_i, \deg f$.

The result is true in this case (we can bound the degree of monomials needed), but does it follow from a more general fact about finite generation of linear monoids? Or is this a very special case, with exactly one coefficient of opposite sign than the others?


I think you may be interested in Corollary 2.11 (a) of Bruns, Gubeladze "Polytopes, rings and K-theory":

Let $M$ and $N$ be affine submonoids of $\mathbb{R}^{d}$. Then $M \cap N$ is an affine monoid.

Their definition (Definition 2.1) of an affine monoid is a monoid which is "finitely generated and isomorphic to a submonoid of a free abelian group $\mathbb{Z}^{d}$ for some $d \ge 0$".

See also Orest Bucicovschi's answer to this question which gives a simpler argument for a special case:

For any subgroup $K \subset \mathbb{Z}^{d}$, the monoid $K \cap \mathbb{N}^{d}$ is finitely generated.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.