We call a polynomial $P(x)=a_dx^d+...+a_0$ of degree $d$ nice if $$\frac{2021(|a_d|+...+|a_0|)}{2022}<max_{0 \le i \le d}|a_i|$$Initially Shayan has a sequence of $d$ distinct real numbers; $r_1,...,r_d \neq \pm 1$. At each step he choose a positive integer $N>1$ and raises the $d$ numbers he has to the exponent of $N$, then delete the previous $d$ numbers and constructs a monic polynomial of degree $d$ with these number as roots, then examine whether it is nice or not. Prove that after some steps, all the polynomials that shayan produces would be nice polynomials Proposed by Navid Safaei
Problem
Source: Iranian RMM TST 2021 Day2 P3
Tags: algebra, polynomial
16.04.2021 09:29
BOBTHEGR8 wrote:
sorry if i was unclear. the new $d$ numbers are the roots of the polynomial
19.04.2021 11:52
Solved with p_square This seems very silly for a TST 3 problem(all you say is some term is very large) but here goes- In case there is some $r_i$ such that $|r_i|>1$- WLOG, let $|r_1|\ge \cdots \ge |r_k|>1$ and $1>r_{k+1}\ge \cdots |r_d|\ge 0$. Now, for all large enough $n$, we have $|r_k|^n>10000^d$ and $|r_{k+1}|^n<10000^d$. Now, $|r_1\cdots r_k|^n>10000^d(|P|)$ where $P$ is the product of any arbitrary subset of $\{r_1,\cdots r_d\}$ and not equal to $r_1\cdots r_k$. Thus, we want to show that for these $n$, the polynomial is nice. We want $|a_{d-k}|>2021(|a_0|+\cdots |a_d|-a_{d-k})$. Observe that $|a_{d-k}|>|r_1\cdots r_k|^n-\frac{2^d}{10000^d}(|r_1\cdots r_k|^n)>\frac{|r_1\cdots r_k|^n}{2}$. Now, $2021(|a_0|+|a_1|+\cdots |a_d|-a_{d-k})<2021(\frac{2^d}{10000^d}(|r_1\cdots r_k|^n))<2021\frac{|r_1\cdots r_k|^n}{5000}<\frac{|r_1\cdots r_k|^n}{2}$ and we are done. Now, if no $|r_i|>1$ then for large enough $n$, all $a_i$ are abitrarily small and $a_d=1$ and we are done.
27.05.2021 18:56
Mr.C wrote: We call a polynomial $P(x)=a_dx^d+...+a_0$ of degree $d$ nice if $$\frac{2021(|a_d|+...+|a_0|)}{2022}<max_{0 \le i \le d}|a_i|$$Initially Shayan has a sequence of $d$ distinct real numbers; $r_1,...,r_d \neq \pm 1$. At each step he choose a positive integer $N>1$ and raises the $d$ numbers he has to the exponent of $N$, then delete the previous $d$ numbers and constructs a monic polynomial of degree $d$ with these number as roots, then examine whether it is nice or not. Prove that after some steps, all the polynomials that shayan produces would be nice polynomials Proposed by Navid Safaei I'm confused understanding the problem. So when $r_1 , ... ,r_d$ provided at the first moment, then after steps these $d$ numbers become just $r_1^M , ... ,r_d^M$ for some $M$? I hope the new $d$ numbers are new coefficient(except for 1) which would make the problem more exciting.