arXiv subjects

Jonas Jankauskas

Publications and source records attributed to Jonas Jankauskas.

13 recordsLinked to original sources

arXiv2026

No three algebraic conjugates of degree sixteen sum to zero

Let $d$ be the smallest positive integer, not a multiple of $3$, for which there exists an algebraic number $\al$ of degree $d$ over $\mathbb{Q}$ whose three algebraic conjugates add to zero. We prove that $d=20$. This is derived from the following result: for any linear relation $\sum_{j=1}^d a_j \al_j=0$ with coefficients $a_j\in\mathbb{Z}$ among the conjugates $\al_j$ of an algebraic number of degree $d=p^m$, where $p$ is a prime number, $m \geq 1$, the sum $\sum_{j=1}a_j$ is divisible by $p$. If $d=2p^m$, $p\geq 3$ and $\sum_{j=1}^d|a_d| < p$, then $\sum_{j=1}a_j$ is an even number.

math.NT↗

arXiv2026

Linear relations of four conjugates of an algebraic number of degree eight

We characterize all algebraic numbers $α$ of degree $8$ for which there exist four distinct algebraic conjugates $α_1$, $α_2$, $α_3$, $α_4$ of $α$ satisfying the linear relation $α_{1}=α_{2}+α_{3}+α_{4}$. Analogous characterization is obtained for the linear relation $α_{1}+α_{2}+α_{3}+α_{4}=0$. In particular, when an algebraic number $α$ of degree $8$ has a non-even minimal polynomial and possesses exactly six distinct linear relations of the form $α_{i_1}+α_{i_2}+α_{i_3}+α_{i_4}=0$, we prove that $α$ is a sum of a quadratic and a quartic algebraic number.

math.NT↗

arXiv2026

No three algebraic conjugates of degree sixteen sum to zero

Let $d$ be the smallest positive integer, not divisible by $3$, for which there exists an algebraic number over $\mathbb{Q}$ of degree $d$ whose some three algebraic conjugates sum to zero. Employing the classification of vertex-transitive graphs on 16 vertices of degree 6, we prove that $d\neq 16$. This, combined with results obtained by Dubickas, Smyth and Stong \cite{DubickasSmyth2006}, Dubickas and Jankauskas \cite{DubickasJankauskas2015} and Virbalas \cite{Virbalas2025a}, implies that $d=20$.

math.NT↗

arXiv2022

On $\mathbf{2\times2}$ determinants originating from survival probabilities in homogeneous discrete time risk model

We analyze $2\times 2$ Hankel-like determinants $D_n$ that arise in the initial values problem for the ultimate time survival probability $φ(u)$ in a homogeneous discrete time risk model $W(n)=u+κn+\sum_{i=1}^nZ_i$, where $Z_i$ are positive integer valued i.i.d. random claims, the initial surplus $u \in \mathbb{N}_0$ and the income rate $κ=2$. We prove the asymptotic version of a recent conjecture on the non--vanishing and monotonicity of $D_n$ and derive explicit formulas for the initial values $φ(0)$, $φ(1)$ of a recurrence that yields survival probabilities. In cases when $Z_i$ are Bernoulli or Geometrically distributed, the conjecture on $D_n$ is shown to hold for all $n\in\mathbb{N}_0$. Additionally, a generating function $Ξ(s)$ for ultimate survival probabilities $φ(u)$ is derived.

math.PR↗

arXiv2022

Multi seasonal discrete time risk model revisited

In this work we set up the distribution function of $\mathcal{M}:=\sup_{n\geqslant1}\sum_{i=1}^{n}{(Z_i-1)}$, where the random walk $\sum_{i=1}^{n}Z_i, n\in\mathbb{N},$ is generated by $N$ periodically occurring distributions and the integer-valued and non-negative random variables $Z_1,\,Z_2,\,\ldots$ are independent. The considered random walk generates so-called multi seasonal discrete time risk model, and a known distribution of random variable $\mathcal{M}$ enables to calculate ultimate time ruin or survival probability. Verifying obtained theoretical statements we demonstrate several computational examples for survival probability $\mathbb{P}(\mathcal{M}< u)$ when $N=2,\,3$ or $10$.

math.PR↗

arXiv2021

Rational matrix digit systems

Let $A$ be a $d \times d$ matrix with rational entries which has no eigenvalue $λ\in \mathbb{C}$ of absolute value $|λ| < 1$ and let $\mathbb{Z}^d[A]$ be the smallest nontrivial $A$-invariant $\mathbb{Z}$-module. We lay down a theoretical framework for the construction of digit systems $(A, \mathcal{D})$, where $\mathcal{D}\subset \mathbb{Z}^d[A]$ finite, that admit finite expansions of the form \[ \mathbf{x}= \mathbf{d}_0 + A \mathbf{d}_1 + \cdots + A^{\ell-1}\mathbf{d}_{\ell-1} \qquad(\ell\in \mathbb{N},\;\mathbf{d}_0,\ldots,\mathbf{d}_{\ell-1} \in \mathcal{D}) \] for every element $\mathbf{x}\in \mathbb{Z}^d[A]$. We put special emphasis on the explicit computation of small digit sets $\mathcal{D}$ that admit this property for a given matrix $A$, using techniques from matrix theory, convex geometry, and the Smith Normal Form. Moreover, we provide a new proof of general results on this finiteness property and recover analogous finiteness results for digit systems in number fields a unified way.

math.NT↗

arXiv2019

On Newman and Littlewood polynomials with prescribed number of zeros inside the unit disk

We study $\{0, 1\}$ and $\{-1, 1\}$ polynomials $f(z)$, called Newman and Littlewood polynomials, that have a prescribed number $N(f)$ of zeros in the open unit disk $\mathcal{D} = \{z \in \mathbb{C}: |z| < 1\}$. For every pair $(k, n) \in \mathbb{N}^2$, where $n \geq 7$ and $k \in [3, n-3]$, we prove that it is possible to find a $\{0, 1\}$--polynomial $f(z)$ of degree $\text{deg }{f}=n$ with non--zero constant term $f(0) \ne 0$, such that $N(f)=k$ and $f(z) \ne 0$ on the unit circle $\partial\mathcal{D}$. On the way to this goal, we answer a question of D.~W.~Boyd from 1986 on the smallest degree Newman polynomial that satisfies $|f(z)| > 2$ on the unit circle $\partial \mathcal{D}$. This polynomial is of degree $38$ and we use this special polynomial in our constructions. We also identify (without a proof) all exceptional $(k, n)$ with $k \in \{1, 2, 3, n-3, n-2, n-1\}$, for which no such $\{0, 1\}$--polynomial of degree $n$ exists: such pairs are related to regular (real and complex) Pisot numbers. Similar, but less complete results for $\{-1, 1\}$ polynomials are established. We also look at the products of spaced Newman polynomials and consider the rotated large Littlewood polynomials. Lastly, based on our data, we formulate a natural conjecture about the statistical distribution of $N(f)$ in the set of Newman and Littlewood polynomials.

math.NT↗

arXiv2019

Linear relations with conjugates of a Salem number

In this paper we consider linear relations with conjugates of a Salem number $α$. We show that every such a relation arises from a linear relation between conjugates of the corresponding totally real algebraic integer $α+1/α$. It is also shown that the smallest degree of a Salem number with a nontrivial relation between its conjugates is $8$, whereas the smallest length of a nontrivial linear relation between the conjugates of a Salem number is $6$.

math.NT↗

arXiv2018

Characterization of rational matrices that admit finite digit representations

Let $A$ be an $n \times n$ matrix with rational entries and let \[ \mathbb{Z}^n[A] := \bigcup_{k=1}^{\infty} \left( \mathbb{Z}^n + A\mathbb{Z}^n + \dots + A^{k-1}\mathbb{Z}^n\right) \] be the minimal $A$-invariant $\mathbb{Z}$-module containing the lattice $\mathbb{Z}^n$. If $\mathcal{D}\subset\mathbb{Z}^n[A]$ is a finite set we call the pair $(A,\mathcal{D})$ a digit system. We say that $(A,\mathcal{D})$ has the finiteness property if each $\mathbf{z} \in \mathbb{Z}^n[A]$ can be written in the form \[ \mathbf{z} = \mathbf{d}_0 + A\mathbf{d}_1 + \dots + A^k\mathbf{d}_k, \] with $k\in\mathbb{N}$ and digits $\mathbf{d}_j \in \mathcal{D}$ for $0\le j\le k$. We prove that for a given matrix $A \in M_n(\mathbb{Q})$ there is a finite set $\mathcal{D}\subset\mathbb{Z}^n[A]$ such that $(A, \mathcal{D})$ has the finiteness property if and only if $A$ has no eigenvalue of absolute value $< 1$. This result is the matrix analogue of the height reducing property of algebraic numbers. In proving this result we also characterize integer polynomials $P \in \mathbb{Z}[x]$ that admit digit systems having the finiteness property in the quotient ring $\mathbb{Z}[x]/(P)$.

math.NT↗

arXiv2016

On Newman and Littlewood multiples of Borwein polynomials

A Newman polynomial has all the coefficients in $\{ 0,1\}$ and constant term 1, whereas a Littlewood polynomial has all coefficients in $\{-1,1\}$. We call $P(X)\in\mathbb{Z}[X]$ a Borwein polynomial if all its coefficients belong to $\{ -1,0,1\}$ and $P(0)\neq 0$. By exploiting an algorithm which decides whether a given monic integer polynomial with no roots on the unit circle $|z|=1$ has a non-zero multiple in $\mathbb{Z}[X]$ with coefficients in a finite set $\mathcal{D} \subset \mathbb{Z}$, for every Borwein polynomial of degree at most 9 we determine whether it divides any Littlewood or Newman polynomial. In particular, we show that every Borwein polynomial of degree at most 8 which divides some Newman polynomial divides some Littlewood polynomial as well. In addition to this, for every Newman polynomial of degree at most 11, we check whether it has a Littlewood multiple, extending the previous results of Borwein, Hare, Mossinghoff, Dubickas and Jankauskas.

math.NT↗

arXiv2014

There are no two non-real conjugates of a Pisot number with the same imaginary part

We show that the number $α=(1+\sqrt{3+2\sqrt{5}})/2$ with minimal polynomial $x^4-2x^3+x-1$ is the only Pisot number whose four distinct conjugates $α_1,α_2,α_3,α_4$ satisfy the additive relation $α_1+α_2=α_3+α_4$. This implies that there exists no two non-real conjugates of a Pisot number with the same imaginary part and also that at most two conjugates of a Pisot number can have the same real part. On the other hand, we prove that similar four term equations $α_1 = α_2 + α_3+α_4$ or $α_1 + α_2 + α_3 + α_4 =0$ cannot be solved in conjugates of a Pisot number $α$. We also show that the roots of the Siegel's polynomial $x^3-x-1$ are the only solutions to the three term equation $α_1+α_2+α_3=0$ in conjugates of a Pisot number. Finally, we prove that there exists no Pisot number whose conjugates satisfy the relation $α_1=α_2+α_3$.

math.NT↗

arXiv2014

The $t$-metric Mahler measures of surds and rational numbers

A. Dubickas and C. Smyth introduced the metric Mahler measure $$ M_1(α) = \inf\left\{\sum_{n=1}^N M(α_n): N \in \mathbb N, α_1 \cdots α_N = α\right\}, $$ where $M(α)$ denotes the usual (logarithmic) Mahler measure of $α\in \overline{\mathbb Q}$. This definition extends in a natural way to the $t$-metric Mahler measure by replacing the sum with the usual $L_t$ norm of the vector $(M(α_1), \dots, M(α_N))$ for any $t\geq 1$. For $α\in \mathbb Q$, we prove that the infimum in $M_t(α)$ may be attained using only rational points, establishing an earlier conjecture of the second author. We show that the natural analogue of this result fails for general $α\in\overline{\mathbb Q}$ by giving an infinite family of quadratic counterexamples. As part of this construction, we provide an explicit formula to compute $M_t(D^{1/k})$ for a square-free $D \in \mathbb N$.

math.NT↗

arXiv2012

On the equation $f(g(x)) =f(x)h^m(x)$ for composite polynomials

In this paper we solve the equation $f(g(x))=f(x)h^m(x)$ where $f(x)$, $g(x)$ and $h(x)$ are unknown polynomials with coefficients in an arbitrary field $K$, $f(x)$ is non-constant and separable, $°g \geq 2$, the polynomial $g(x)$ has non-zero derivative $g'(x) \ne 0$ in $K[x]$ and the integer $m \geq 2$ is not divisible by the characteristic of the field $K$. We prove that this equation has no solutions if $°f \geq 3$. If $°f = 2$, we prove that $m = 2$ and give all solutions explicitly in terms of Chebyshev polynomials. The diophantine applications for such polynomials $f(x)$, $g(x)$, $h(x)$ with coefficients in $\Q$ or $\Z$ are considered in the context of the conjecture of Cassaign et. al on the values of Louiville's $λ$ function at points $f(r)$, $r \in \Q$.

math.NT↗