Given a + b + c = 1 and a , b , c > 0 , find the minimum value of a 2 + 2 b 2 + c 2 .
This section requires Javascript.
You are seeing this because something didn't load right. We suggest you, (a) try
refreshing the page, (b) enabling javascript if it is disabled on your browser and,
finally, (c)
loading the
non-javascript version of this page
. We're sorry about the hassle.
Used Titu"s Lemma,
d a 2 + e b 2 + f f 2 ≥ d + e + f ( a + b + c ) 2
= a 2 + 2 1 b 2 + c 2 ≥ 1 + 2 1 + 1 ( a + b + c ) 2
minimum value = 5 2 = 0 . 4
Log in to reply
GREAT GENERALIZATION.!!!
Nice method.... Better than cuachi shwarz
I also solved this problem using Cauchy's inequality. Awesome solution !
Best solution!!
THE ANSWER TO THE QUESTION IS WRONG AS EQUALITY CAN BE ACHIEVED AT SOME LOW VALUE ALSO. IF a=c=2^.5/(2 2^.5 + 1) and b=1/(2 2^.5 + 1) hence min value is achieved when a=c=2^.5b.hence the correct answer should be 0.3411
Log in to reply
Substituting this into the original expression yields 0 . 4 0 9 4 . You may have forgotten about the coefficient of 2 in front of b 2 . Ignoring this coefficient does indeed give 0 . 3 4 1 1 .
Did the same way
vinay agrawal Sir , please check your calculation, if you put all those values, we get an answer greater than 0.4
Vinay.............................................................
Using Cauchy-Schwartz inequality he have:
( 1 2 + ( 2 1 ) 2 + ( 2 1 ) 2 + 1 2 ) ( a 2 + b 2 + b 2 + c 2 ) ≥ ( a + 2 b + 2 b + c ) 2
⇒ 2 5 ( a 2 + 2 b 2 + c 2 ) ≥ ( a + b + c ) 2 = 1
⇒ a 2 + 2 b 2 + c 2 ≥ 5 2
Sir, the third term in the first line should be 2 1 2 . Great solution otherwise, as always!
For sake of variety, how about some Lagrange multipliers? We wish to minimize f ( a , b , c ) = a 2 + 2 b 2 + c 2 given the restraint g ( a , b , c ) = a + b + c − 1 = 0 .
Then the Lagrange equations ∂ ( a , b , c ) ∂ f ( a , b , c ) = λ ∂ ( a , b , c ) ∂ g ( a , b , c ) are
2 a = λ , 4 b = λ , 2 c = λ ⟹ a = 2 λ , b = 4 λ , c = 2 λ .
Plugging these into the constraint gives us that 4 5 λ = 1 ⟹ λ = 5 4 .
Thus a = c = 5 2 and b = 5 1 , giving us
f ( 5 2 , 5 1 , 5 2 ) = 2 5 4 + 2 ∗ 2 5 1 + 2 5 4 = 2 5 1 0 = 5 2 = 0 . 4 as the minimum.
(Note that, for example, if a = b = c = 3 1 we would have a 2 + 2 b 2 + c 2 = 9 4 > 5 2 , so we can conclude that the solution found above is indeed the minimum. Also note that by this method we can drop the restriction that a , b , c be positive.)
Can you please elaborately describe how and why your method works, or give a link where i can learn it, or make a wiki about it?
Log in to reply
I've always found that this site provides very clear overviews of topics in mathematics. The method of Lagrange multipliers is usually dealt with in 1st year university because the theory behind it is somewhat involved, (partial derivatives, gradients, etc.), but using the method is quite straightforward. I posted my solution just to introduce Lagrange multipliers to members who weren't already familiar with this very useful approach. :)
This video provides the method of Lagrange Multipliers, but not the explanation.
However, I want to say that using Lagrange Multipliers in this problem is not the best choice. I always leave it as my final move.
Log in to reply
Yes, I normally use Lagrange multipliers, (and L'Hopital's rule with limits), as last resorts. I just wanted to introduce the method into the discussion in case others weren't familiar with it. :)
That was how I solved it!
ANSWER IS CUBE ROOT OF 2
Make sure you check your critical point against the Hessian Matrix of f to see if it's positive-definite (i.e. yields a global minimum)!
It is symmetric about a and c , Thus putting a=c=p
we have 2p+b=1
and
2 p 2 + 2 b 2 = 1
solving above equations yield the quadratic 1 0 p 2 − 8 p + 2
which gives a minimum at p=0.4 which is 0.4 itself
I'm sorry but the above ecuation system doesn't give that resultant quadratic.
Applying Cauchy-Schwarz's Inequality, we have ( a 2 + 2 b 2 + c 2 ) ( 1 + 2 1 + 1 ) ≥ ( a + b + c ) 2 = 1 Thus, ( a 2 + 2 b 2 + c 2 ) ≥ 1 + 2 1 + 1 1 = 5 2 Therefore the minimum possible value is 5 2 = 0 . 4
A physicist's full proof:
We can see that b carries a lot more weight in the final sum a 2 + 2 b 2 + c 2 . Seeing this and also that a and c must be symmetric, we can guess that the minimum value of the second sum happens when a and c are equal, b is half the value of those two, and their sum is equal to 1 . This happens when
a = c = 5 2
b = 5 1
Which would give us a 2 + 2 b 2 + c 2 = 2 5 1 0 = 0 . 4 .
Let k = a + c = 1 − b , ie. c = k − a , b = 1 − k
a 2 + 2 b 2 + c 2 = a 2 + c 2 + 2 b 2 = a 2 + ( k − a ) 2 + 2 ( 1 − k ) 2 = 2 a 2 − 2 k a + a 2 + 2 k 2 − 4 k + 2 = 2 ( a − 2 1 k ) 2 + 2 5 ( k − 5 4 ) 2 + 5 2
Hence, when c = a = 2 1 k , k = 5 4 , ie. b = 1 − 5 4 = 5 1
a 2 + 2 b 2 + c 2 has minimum value of 5 2
as someone mentioned, I also figured out that this would be symmetric about a and c.
since a+b+c =1 and a^2 + 2 b^2 + c^2 both are having equal weight of a and c, I assumed a=c=k
now
it's a simple differentiation problem given 2a+b =1 minimize 2 a^2 + 2 b^2, simply by taking derivative of b w.r.t a.
Used Titu's Lemma where
a
1
=
a
,
a
2
=
b
,
a
3
=
c
,
b
1
=
1
,
b
2
=
2
1
and
b
3
=
1
Computing is generally applicable for many types of questions:
From 0.4444 to 0.4, answer is 0.4 with a = 2/ 5, b = 1/ 5 and c = 2/ 5.
PROGRAM Search;
USES CRT;
VAR a, b, c: LONGINT; m, y: EXTENDED;
BEGIN CLRSCR; m:=4/ 9;
FOR a:=1 TO 1000 DO
FOR b:=1 TO 1000 DO
BEGIN
c:=1000-a-b;
y:=SQR(0.001*a)+2.0*SQR(0.001*b)+SQR(0.001*c);
IF (c>0) AND (y<m) THEN
BEGIN
m:=y;
WRITELN(m:1:4, ' ', 0.001*a:1:2, ' ', 0.001*b:1:2, ' ', 0.001*c:1:2);
END
END
END.
Since a = c ought to be true after intuition from above,
a + b + a = 1 and a^2 + 2 b^2 + a^2 need for a minimum,
2 a^2 + 2 (1 - 2 a)^2 = 2 [5 (a - 2/ 5 )^2 + 1/ 5] means a = c = 2/ 5 while b = 1 - 4/ 5 = 1/ 5 for 2 [1/ 5] = 0.4 {Answer}
We can see that (b) carries a lot more weight in the final sum a^2 + 2 b^2 + c^2 .
Seeing this and also that (a) and (c) must be symmetric, we can guess that the minimum value of the second sum happens when (a) and (c) are equal, (b) is half the value of those two, and their sum is equal to 1 . This happens when:
a = c = 2/5
b = 1/5
Which would give us a^2 + 2 b^2 + c^2 = (2/5)^2 + 2 (1/5)^2 + (2/5)^2 = 4/10
= 2/5 = 0.4
Well i'm not that good with inequalities so I will solve this problem without complicated inequalities or theorems
First, according to QM - AM
a
2
+
c
2
obtain minimum at
a
=
c
We get 2 a + b = 1 and want to minimize 2 a 2 + 2 b 2
Hint : find a circle centered at ( 0 , 0 ) with the smallest radius r but still intersect the line 2 a + b = 1
The answer to this problem is the value of 2 r 2
you can prefer the solution using lagrange multiplier
Problem Loading...
Note Loading...
Set Loading...
using Cauchy inequality :
( ( a ) × ( 1 ) + ( 2 b ) × ( 2 1 ) + ( c ) × ( 1 ) ) 2 ≤ ( a 2 + 2 b 2 + c 2 ) ( 1 2 + ( 2 1 ) 2 + 1 2 ) ( a 2 + 2 b 2 + c 2 ) m i n = 2 / 5 = 0 . 4 .