Difference between revisions of "2000 USAMO Problems/Problem 6"
Fibonacci97 (talk | contribs) (Created page with "==Problem== ''[Insert USAMO 2000 Problem 6 here]'' ==Solution== ''[Insert complete solution(s) to USAMO 2000 Problem 6 here]''") |
(→Problem) (Tag: Replaced) |
||
(5 intermediate revisions by 4 users not shown) | |||
Line 1: | Line 1: | ||
==Problem== | ==Problem== | ||
− | + | Let <math>a_1, b_1, a_2, b_2, \dots , a_n, b_n</math> be nonnegative real numbers. Prove that | |
+ | <cmath>\sum_{i, j = 1}^{n} \min\{a_ia_j, b_ib_j\} \le \sum_{i, j = 1}^{n} \min\{a_ib_j, a_jb_i\}.</cmath> | ||
==Solution== | ==Solution== | ||
− | '' | + | ''Credit for this solution goes to Ravi Boppana.'' |
+ | |||
+ | '''Lemma 1:''' If <math>r_1, r_2, \ldots , r_n</math> are non-negative reals and <math>x_1, x_2, \ldots x_n</math> are reals, then | ||
+ | |||
+ | <cmath> \sum_{i, j}\min(r_{i}, r_{j}) x_{i}x_{j}\ge 0. </cmath> | ||
+ | |||
+ | '''Proof:''' Without loss of generality assume that the sequence <math>\{r_i\}</math> is increasing. For convenience, define <math>r_0=0</math>. The LHS of our inequality becomes | ||
+ | |||
+ | <cmath>\sum_{i}r_{i}x_{i}^{2}+2\sum_{i < j}r_{i}x_{i}x_{j}\, .</cmath> | ||
+ | |||
+ | This expression is equivalent to the sum | ||
+ | |||
+ | <cmath>\sum_{i}(r_{i}-r_{i-1})\biggl(\sum_{j=i}^{n}x_{j}\biggr)^{2}\, .</cmath> | ||
+ | |||
+ | Each term in the summation is non-negative, so the sum itself is non-negative. <math>\blacksquare</math> | ||
+ | |||
+ | We now define <math>r_i=\frac{\max(a_i,b_i)}{\min(a_i,b_i)}-1</math>. If <math>\min(a_i,b_i)=0</math>, then let <math>r_i</math> be any non-negative number. Define <math>x_i=\text{sgn}(a_i-b_i)\min(a_i,b_i)</math>. | ||
+ | |||
+ | '''Lemma 2:''' <math>\min(a_{i}b_{j}, a_{j}b_{i})-\min(a_{i}a_{j}, b_{i}b_{j}) =\min(r_{i}, r_{j}) x_{i}x_{j}</math> | ||
+ | |||
+ | '''Proof:''' Switching the signs of <math>a_i</math> and <math>b_i</math> preserves inequality, so we may assume that <math>a_i>b_i</math>. Similarly, we can assume that <math>a_j>b_j</math>. If <math>b_ib_j=0</math>, then both sides are zero, so we may assume that <math>b_i</math> and <math>b_j</math> are positive. We then have from the definitions of <math>r_i</math> and <math>x_i</math> that | ||
+ | |||
+ | <cmath>\begin{eqnarray*}r_{i}& = &\frac{a_{i}}{b_{i}}-1\\ r_{j}& = &\frac{a_{j}}{b_{j}}-1\\ x_{i}& = & b_{i}\\ x_{j}& = & b_{j}\, .\end{eqnarray*}</cmath> | ||
+ | |||
+ | This means that | ||
+ | |||
+ | <cmath>\begin{eqnarray*}\min(r_{i}, r_{j}) x_{i}x_{j}& = &\min\bigl(\frac{a_{i}}{b_{i}}-1,\frac{a_{j}}{b_{j}}-1\bigr) b_{i}b_{j}\\ & = &\min(a_{i}b_{j}, a_{j}b_{i})-b_{i}b_{j}\\ & = &\min(a_{i}b_{j}, a_{j}b_{i})-\min(a_{i}a_{j}, b_{i}b_{j})\, .\end{eqnarray*}</cmath> | ||
+ | |||
+ | This concludes the proof of Lemma 2. <math>\blacksquare</math> | ||
+ | |||
+ | We can then apply Lemma 2 and Lemma 1 in order to get that | ||
+ | |||
+ | <cmath>\begin{eqnarray*}\sum_{i,j}\min(a_{i}b_{j}, a_{j}b_{i})-\sum_{i, j}\min(a_{i}a_{j}, b_{i}b_{j}) & = &\sum_{i, j}\left[\min(a_{i}b_{j}, a_{j}b_{i})-\min(a_{i}a_{j}, b_{i}b_{j})\right]\\ & = &\sum_{i, j}\min(r_{i}, r_{j}) x_{i}x_{j}\\ &\ge & 0\, .\end{eqnarray*}</cmath> | ||
+ | |||
+ | This implies the desired inequality. | ||
+ | |||
+ | == See Also == | ||
+ | {{USAMO newbox|year=2000|num-b=5|after=Last Question}} | ||
+ | |||
+ | [[Category:Olympiad Algebra Problems]] | ||
+ | [[Category:Olympiad Inequality Problems]] | ||
+ | {{MAA Notice}} |
Latest revision as of 14:08, 15 July 2021
Problem
Let be nonnegative real numbers. Prove that
Solution
Credit for this solution goes to Ravi Boppana.
Lemma 1: If are non-negative reals and are reals, then
Proof: Without loss of generality assume that the sequence is increasing. For convenience, define . The LHS of our inequality becomes
This expression is equivalent to the sum
Each term in the summation is non-negative, so the sum itself is non-negative.
We now define . If , then let be any non-negative number. Define .
Lemma 2:
Proof: Switching the signs of and preserves inequality, so we may assume that . Similarly, we can assume that . If , then both sides are zero, so we may assume that and are positive. We then have from the definitions of and that
This means that
This concludes the proof of Lemma 2.
We can then apply Lemma 2 and Lemma 1 in order to get that
This implies the desired inequality.
See Also
2000 USAMO (Problems • Resources) | ||
Preceded by Problem 5 |
Followed by Last Question | |
1 • 2 • 3 • 4 • 5 • 6 | ||
All USAMO Problems and Solutions |
The problems on this page are copyrighted by the Mathematical Association of America's American Mathematics Competitions.