Unofficial archive — problems, solutions & results © IMC, reproduced with permission.

Studolymp / IMC / 1996

IMC 1996
contestants · problems 12 (6+6) · scale 0–10 · per-problem no

No individual results were published for this edition — problems and solutions only.

Problems

Day 1

P1

Let for j=0,,nj = 0, \dots, n, aj=a0+jda_j = a_0 + jd, where a0a_0, dd are fixed real numbers. Put A=(a0a1a2ana1a0a1an1a2a1a0an2anan1an2a0).A = \begin{pmatrix} a_0 & a_1 & a_2 & \dots & a_n \\ a_1 & a_0 & a_1 & \dots & a_{n-1} \\ a_2 & a_1 & a_0 & \dots & a_{n-2} \\ \dots & \dots & \dots & \dots & \dots \\ a_n & a_{n-1} & a_{n-2} & \dots & a_0 \end{pmatrix}. Calculate det(A)\det(A), where det(A)\det(A) denotes the determinant of AA.

P2

Evaluate the definite integral ππsinnx(1+2x)sinxdx,\int_{-\pi}^{\pi} \frac{\sin nx}{(1 + 2^x) \sin x}\,dx, where nn is a natural number.

P3

The linear operator AA on the vector space VV is called an involution if A2=EA^2 = E where EE is the identity operator on VV. Let dimV=n<\dim V = n < \infty.

(i) Prove that for every involution AA on VV there exists a basis of VV consisting of eigenvectors of AA.

(ii) Find the maximal number of distinct pairwise commuting involutions on VV.

P4

Let a1=1a_1 = 1, an=1nk=1n1akanka_n = \dfrac{1}{n} \sum\limits_{k=1}^{n-1} a_k a_{n-k} for n2n \ge 2. Show that

(i) lim supnan1/n<21/2\limsup\limits_{n \to \infty} |a_n|^{1/n} < 2^{-1/2};

(ii) lim supnan1/n2/3\limsup\limits_{n \to \infty} |a_n|^{1/n} \ge 2/3.

P5

(i) Let aa, bb be real numbers such that b0b \le 0 and 1+ax+bx201 + ax + bx^2 \ge 0 for every xx in [0,1][0,1]. Prove that limn+n01(1+ax+bx2)ndx={1aif a<0,+if a0.\lim_{n \to +\infty} n \int_0^1 (1 + ax + bx^2)^n dx = \begin{cases} -\dfrac{1}{a} & \text{if } a < 0, \\ +\infty & \text{if } a \ge 0. \end{cases}

(ii) Let f:[0,1][0,)f : [0,1] \to [0,\infty) be a function with a continuous second derivative and let f(x)0f''(x) \le 0 for every xx in [0,1][0,1]. Suppose that L=limnn01(f(x))ndxL = \lim\limits_{n \to \infty} n \int_0^1 (f(x))^n dx exists and 0<L<+0 < L < +\infty. Prove that ff' has a constant sign and minx[0,1]f(x)=L1\min\limits_{x \in [0,1]} |f'(x)| = L^{-1}.

P6

Upper content of a subset EE of the plane R2\mathbb{R}^2 is defined as C(E)=inf{i=1ndiam(Ei)}\mathcal{C}(E) = \inf \left\{ \sum_{i=1}^{n} \operatorname{diam}(E_i) \right\} where inf\inf is taken over all finite families of sets E1,,EnE_1, \dots, E_n, nNn \in \mathbb{N}, in R2\mathbb{R}^2 such that Ei=1nEiE \subset \bigcup\limits_{i=1}^{n} E_i.

Lower content of EE is defined as K(E)=sup{lenght(L):L is a closed line segmentonto which E can be contracted}.\begin{align*} \mathcal{K}(E) = \sup \{ \operatorname{lenght}(L) :{}& \text{$L$ is a closed line segment} \\ & \text{onto which $E$ can be contracted} \}.

\end{align*} Show that

(a) C(L)=lenght(L)\mathcal{C}(L) = \operatorname{lenght}(L) if LL is a closed line segment;

(b) C(E)K(E)\mathcal{C}(E) \ge \mathcal{K}(E);

(c) the equality in (b) needs not hold even if EE is compact.

Hint. If E=TTE = T \cup T' where TT is the triangle with vertices (2,2)(-2,2), (2,2)(2,2) and (0,4)(0,4), and TT' is its reflexion about the xx-axis, then C(E)=8>K(E)\mathcal{C}(E) = 8 > \mathcal{K}(E).

Remarks: All distances used in this problem are Euclidian. Diameter of a set EE is diam(E)=sup{dist(x,y):x,yE}\operatorname{diam}(E) = \sup \{ \operatorname{dist}(x,y) : x, y \in E \}. Contraction of a set EE to a set FF is a mapping f:EFf : E \mapsto F such that dist(f(x),f(y))dist(x,y)\operatorname{dist}(f(x), f(y)) \le \operatorname{dist}(x,y) for all x,yEx, y \in E. A set EE can be contracted onto a set FF if there is a contraction ff of EE to FF which is onto, i.e., such that f(E)=Ff(E) = F. Triangle is defined as the union of the three segments joining its vertices, i.e., it does not contain the interior.

Day 2

P7

Prove that if f:[0,1][0,1]f : [0,1] \to [0,1] is a continuous function, then the sequence of iterates xn+1=f(xn)x_{n+1} = f(x_n) converges if and only if limn(xn+1xn)=0.\lim_{n \to \infty} (x_{n+1} - x_n) = 0.

P8

Let θ\theta be a positive real number and let cosht=et+et2\cosh t = \dfrac{e^t + e^{-t}}{2} denote the hyperbolic cosine. Show that if kNk \in \mathbb{N} and both coshkθ\cosh k\theta and cosh(k+1)θ\cosh(k+1)\theta are rational, then so is coshθ\cosh \theta.

P9

Let GG be the subgroup of GL2(R)GL_2(\mathbb{R}), generated by AA and BB, where A=[2001],B=[1101].A = \begin{bmatrix} 2 & 0 \\ 0 & 1 \end{bmatrix}, \qquad B = \begin{bmatrix} 1 & 1 \\ 0 & 1 \end{bmatrix}. Let HH consist of those matrices (a11a12a21a22)\begin{pmatrix} a_{11} & a_{12} \\ a_{21} & a_{22} \end{pmatrix} in GG for which a11=a22=1a_{11} = a_{22} = 1.

(a) Show that HH is an abelian subgroup of GG.

(b) Show that HH is not finitely generated.

Remarks. GL2(R)GL_2(\mathbb{R}) denotes, as usual, the group (under matrix multiplication) of all 2×22 \times 2 invertible matrices with real entries (elements). Abelian means commutative. A group is finitely generated if there are a finite number of elements of the group such that every other element of the group can be obtained from these elements using the group operation.

P10

Let BB be a bounded closed convex symmetric (with respect to the origin) set in R2\mathbb{R}^2 with boundary the curve Γ\Gamma. Let BB have the property that the ellipse of maximal area contained in BB is the disc DD of radius 1 centered at the origin with boundary the circle CC. Prove that AΓA \cap \Gamma \ne \emptyset for any arc AA of CC of length l(A)π2l(A) \ge \dfrac{\pi}{2}.

P11

(i) Prove that limx+n=1nx(n2+x)2=12.\lim_{x \to +\infty} \sum_{n=1}^{\infty} \frac{nx}{(n^2 + x)^2} = \frac{1}{2}.

(ii) Prove that there is a positive constant cc such that for every x[1,)x \in [1, \infty) we have n=1nx(n2+x)212cx.\left| \sum_{n=1}^{\infty} \frac{nx}{(n^2 + x)^2} - \frac{1}{2} \right| \le \frac{c}{x}.

P12

(Carleman's inequality)

(i) Prove that for every sequence {an}n=1\{a_n\}_{n=1}^{\infty}, such that an>0a_n > 0, n=1,2,n = 1, 2, \dots and n=1an<\sum\limits_{n=1}^{\infty} a_n < \infty, we have n=1(a1a2an)1/n<en=1an,\sum_{n=1}^{\infty} (a_1 a_2 \cdots a_n)^{1/n} < e \sum_{n=1}^{\infty} a_n, where ee is the natural log base.

(ii) Prove that for every ε>0\varepsilon > 0 there exists a sequence {an}n=1\{a_n\}_{n=1}^{\infty}, such that an>0a_n > 0, n=1,2,n = 1, 2, \dots, n=1an<\sum\limits_{n=1}^{\infty} a_n < \infty and n=1(a1a2an)1/n>(eε)n=1an.\sum_{n=1}^{\infty} (a_1 a_2 \cdots a_n)^{1/n} > (e - \varepsilon) \sum_{n=1}^{\infty} a_n.