Community

Want to learn how to tackle those tough AMC/AIME/Olympiad algebra problems? Check out Art of Problem Solving's NEW Intermediate Algebra by Richard Rusczyk and Mathew Crawford. Over 1600 problems!
Login Register Memberlist Search AoPS Blogs Contests Galleries Forum Index
The time now is Tue Oct 14, 2008 7:06 am
All times are UTC - 7 (DST in action)
AoPSWiki WoWCircumcircle, AMC 8
View posts since last visit
View unanswered posts
Sum (discrete) inequality
Moderators: Arne, blahblahblah, Cezar Lupu, darij grinberg, harazi, Megus, N.T.TUAN, orl, pbornsztein, pvthuan
Post new topic   Reply to topic View previous topicView next topic
3 Posts • Page 1 of 1
Author Message
flip2004
Yang-Mills Theory
Yang-Mills Theory

Offline
Joined: 18 Jul 2004
Posts: 917
Location: Sibiu
Romania
 
#1
Sum (discrete) inequality
ALEX collection

Let f(x)=\displaystyle Ax^3+Bx^2+CX+D with real coefficients. Then
\sum\limits_{j=1}^{n}f(j) \le n^2(8n^2-7)\cdot\max\limits_{x\in [0,1]}|f(x)|\; \; \; ,\; \; \; \; n=1,2,...\; .
Prove or disprove that this inequality cannot be improved.

_________________
Alex. Lupas
Last edited by flip2004 on Thu Jan 12, 2006 7:04 am; edited 1 time in total 
PostPosted: Thu Sep 16, 2004 12:15 am  Back to top 
  ProfilePM
harazi
Birch & Swinnerton Dyer
Birch & Swinnerton Dyer

Offline
Joined: 12 Nov 2003
Posts: 5405
Location: Paris
RomaniaFrance
 
#2
No, I think the inequality cannot be improved. For this take the polynomial ( guess how I found it). And after many many computations I think we get equality in the inequality. The problem is really difficult and nice, but I solved it immediately. Do you remember that problem I asked you, flip 2004? And I still do not know your answer. The main thing is here to consider the polynomial and to suppose without loss of generaliy that the maximal value of the absolute value of g on [-1,1] is 1. Then ( the main step) is to prove using Lagrange technique that for all x having modulus at least 1 we have that the absolute value of g(x) is at most the absolute value of and after that to use this inequality and sevral computations to finish the proof.

PostPosted: Thu Sep 16, 2004 11:41 am  Back to top 
  ProfilePM
flip2004
Yang-Mills Theory
Yang-Mills Theory

Offline
Joined: 18 Jul 2004
Posts: 917
Location: Sibiu
Romania
 
#3
Nice remarks by Harazi an a possible generalization

Hi Harazi, nice remarks. Indeed, let us try to extend my proposed question:
Denote by the Chebychev polynomial of degree , i.e. T_n(z)=\cos{(n\cdot \arccos{z})} when Let be a linear space of real functions defined on with . Suppose that the linear space of all real polynomials of degree is a subspace of

Generalization of proposed question: If is a linear positive functional , then for any polynomial of degree we have


where \; \displaystyle ||h||=\max\limits_{x\in [0,1]}|h(x)|\;  \; and \; \; \displaystyle T_n^*(x)= T_n(2x-1)\; .


In the proposed question Y =\left\{ f:[1,n]\to {\mathbb R} and A[f]=\sum\limits_{k=1}^{n}f(k) , as observed by Harazi.

Try to consider a linear functional having images
A[f]=\frac{\Gamma(p+q)}{\Gamma(p)\Gamma(q)}\int\limits_0^1 t^{p-1}(1-t)^{q-1}
f\left(1+at \right )\; dt \; \; \; ,\; \; \; \; a>0 , p>0,q>0 .

PostPosted: Sat Sep 18, 2004 11:53 pm  Back to top 
  ProfilePM
Display posts from previous:   Sort by:   
3 Posts • Page 1 of 1
Post new topic   Reply to topic View previous topicView next topic
Jump to:  

You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
You cannot attach files in this forum
You can download files in this forum
You cannot post calendar events in this forum


© Copyright 2008 AoPS Incorporated. All Rights Reserved. • FoundationPrivacyContact Us