\spacedallcaps Probabilistic Effectivity in the Subspace Theorem
\spacedlowsmallcaps Faustin ADICEAM & \spacedlowsmallcaps Victor SHIRANDAMI · 2024
Read \spacedallcaps Probabilistic Effectivity in the Subspace Theorem by \spacedlowsmallcaps Faustin ADICEAM & \spacedlowsmallcaps Victor SHIRANDAMI (2024) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2411-01247.txt.
Description
A 2024 arXiv research paper by \spacedlowsmallcaps Faustin ADICEAM & \spacedlowsmallcaps Victor SHIRANDAMI. The Subspace Theorem due to Schmidt (1972) is a broad generalisation of Roth’s Theorem in Diophantine Approximation (1955) which, in the same way as the latter, suffers a notorious lack of effectivity. This problem is tackled from a probabilistic standpoint by determining the proportion of algebraic linear forms of bounded heights and degrees for which there exists a solution to the Subspace Inequality lying in a subspace of large height.
Catalog record
| CREATOR | \spacedlowsmallcaps Faustin ADICEAM & \spacedlowsmallcaps Victor SHIRANDAMI |
|---|---|
| YEAR | 2024 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 109 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2411-01247
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
\spacedallcaps Probabilistic Effectivity in the Subspace Theorem \spacedlowsmallcaps Faustin ADICEAM & \spacedlowsmallcaps Victor SHIRANDAMI Abstract The Subspace Theorem due to Schmidt (1972) is a broad generalisation of Roth’s Theorem in Diophantine Approximation (1955) which, in the same way as the latter, suffers a notorious lack of effectivity. This problem is tackled from a probabilistic standpoint by determining the proportion of algebraic linear forms of bounded heights and degrees for which there exists a solution to the Subspace Inequality lying in a subspace of large height. The estimates are established for a class of height functions emerging from an analytic parametrisation of the projective space. They are pertinent in the regime where the heights of the algebraic quantities are larger than those of the rational solutions to the inequality under consideration, and are valid for approximation functions more general than the power functions intervening in the original Subspace Theorem. These estimates are further refined in the case of Roth’s Theorem so as to yield a Khintchin–type density version of the so–called Waldschmidt conjecture (which is known to fail pointwise). This answers a question raised by Beresnevich, Bernik and Dodson (2009). † † footnotetext: 1 Laboratoire d’analyse et de mathématiques appliquées (LAMA), Université Paris-Est Créteil, France, [email protected] † † footnotetext: 2 Department of Mathematics, The University of Manchester, United-Kingdom, [email protected] . The support of the Heilbronn Institute of Mathematical Research through the UKRI grant: Additional Funding Programme for Mathematical Sciences (EP/V521917/1) is gratefully acknowledged. 1 Introduction 1.1 Effectivity in the Subspace Theorem. Given an integer n\geq 2 and vectors \bm{x},\bm{y}\in\mathbb{R}^{n} , denote by \bm{x\cdot y} the scalar product between \bm{x} and \bm{y} . In its initial version, the celebrated Subspace Theorem due to Schmidt [ 18 ] states the following : Theorem 1.1 (The Subspace Theorem — Schmidt, 1972) . Let \bm{\alpha}_{1},\dots,\bm{\alpha}_{n}\in\mathbb{R}^{n} be n linearly independant vectors with algebraic coordinates. Fix \varepsilon>0 . Then, the set of solutions in \bm{q}\in\mathbb{Z}^{n}\backslash\{\bm{0}\} to the inequality \prod_{i=1}^{n}\left|\bm{q}\bm{\cdot}\bm{\alpha}_{i}\right|\;\leq\;\left\|\bm{q}\right\|^{-\varepsilon} (1.1) lies in a finite union of proper rational subspaces of \mathbb{Q}^{n} . Generalisations including non–Archimedean absolute values are known — see [ 5 ] for further details. An upper bound for the number of rational subspaces in this statement can be effectively determined — see, e.g., [ 19 ] and the related references. One of the most fundamental open problems in Diophantine Analysis is to establish an effective version of this statement; namely, to determine the maximal height attained by the subspaces in the finite union. The overarching goal undertaken here is to tackle this effectivity problem from a probabilistic point of view by estimating the proportion of algebraic vectors for which the Subspace Inequality admits a large solution; that is, a solution lying in a subspace with large height. Given an integers Q\geq 1 , this probabilistic approach can be very broadly described as such : ( Main Problem — General Formulation ) What is the proportion of the set of n –tuples of algebraic vectors in \mathbb{R}^{n} of fixed degrees such that the Subspace Inequality ( 1.1 ) admits a nonzero integer solution \bm{q} lying in a rational subspace of height at least Q ? A number of terms needs to be specified in this question — this will be the purpose of the rest of the introduction culminating with the statement of the main Theorems 1.3 and 1.6 . For the time being, the following points should be noted : • the notion of ”proportion of algebraic vectors of fixed degrees” is interpreted in terms of a density function as defined in §§ 1.2 & 1.5 . It relies on the definition of a suitable class of height and degree functions for the set of algebraic vectors, which is introduced in § 1.4 ; • Estimates in terms of density are derived asymptotically from counting problems dealing with finite point sets. To formalise this approach, together with the integer Q=Q_{1} measuring the height of the rational subspace appearing in the statement of the Main Problem, another two parameters are introduced : one, denoted by H\geq 1 , measures the maximal height of the algebraic vectors \bm{\alpha}_{i} of fixed degrees, and another one, denoted by Q_{2} , bounds the height of the rational solutions \bm{q} lying in the subspace. • the Main Problem is solved in a quantitative form so as to be able to answer a question such as the following : How big should the real quantities H,Q_{1} and Q_{2} be chosen under the assumption that H\geq Q_{2}\geq Q_{1} to guarantee that at most, say 5%, of the n –tuples of algebraic vectors \bm{\alpha}_{i}\in\mathbb{R}^{n} of given degrees and of heights at most H yield the existence of a solution \bm{q} to the Subspace Inequality ( 1.1 ) meeting these two properties : (i) \bm{q} has height at most Q_{2} and (ii) \bm{q} lies in some rational subspace of height at least Q_{1} ? Of paramount importance is the interplay between the parameters H (measuring the heights of the algebraic vectors \bm{\alpha}_{i} ) and Q_{2} (measuring the heights of the rational solutions \bm{q} ). The main results provide a solution to this question when the parameter H is larger than Q_{2} : this situation, involving the study of the statistical distribution of algebraic vectors, is what is referred to here as the probabilistic regime . The other relevant alternative, viz. when the parameter Q_{2} is larger than H , can be referred to as the pseudo-deterministic regime and will not be broached here 3 . † † footnotetext: 3 The pseudo-deterministic regime essentially comes down to taking into account all possible rational solutions \bm{q} to the Subspace Inequality lying in some subspace of height at least Q_{1} . Solving the Main Problem in this regime is, in fact, little different from the original problem of effectivity : it reduces to finding an effective bound for the heights of the rational solutions in the Subspace Theorem valid for all but a small and quantified set of exceptional n -tuples of algebraic vectors of bounded height. The particular case where the Subspace Theorem is specialised to Roth’s Theorem in Diophantine Approximation is considered with a special interest in the following section. The above approach is then refined, and the obtained results are more complete and more apparent. 1.2 Probabilistic Estimates in Roth’s Theorem. Roth’s Theorem amounts to considering the case n=2 and to setting \bm{\alpha}_{1}=(1,0) and \bm{\alpha}_{2}=(\alpha,1) in Theorem 1.1 : Theorem 1.2 (Roth, 1955) . Assume that \alpha is an irrational algebraic number. Then, given any \varepsilon>0 , the number of solutions in integers q\geq 1 and p\in\mathbb{Z} to the inequality \left|\alpha-\frac{p}{q}\right|\;\leq\;\frac{\Psi(q)}{q} (1.2) is finite when \Psi(q)=q^{-\varepsilon} . More generally, let \Psi\penalty\ :\mathbb{R}_{+}\mapsto\mathbb{R}_{+} be an approximation function; that is, a nonincreasing function tending to zero at infinity. The Main Problem is considered in the generality where Roth’s inequality ( 1.2 ) is satisfied for any approximation function satisfying a convergence or divergence condition (of which the original case \Psi(x)=x^{-\varepsilon} is a particular example). To this end, given a fixed degree d\geq 2 and an integer H\geq 1 , let \mathbb{A}_{d}(H) be the set of real algebraic numbers of degree d and of (naive) height (defined as the maximum of the absolute values of the coefficients