⎩ ⎪ ⎨ ⎪ ⎧ a 2 + b 2 + c 2 + d 2 = 1 K = a b + a c + a d + b c + b d + 3 c d
Let 4 real numbers a , b , c , and d satisfy the two equations above. If the minimum and maximum values of K are m and M , respectively, what is the value of m + M ?
Give your answer to 3 decimal places.
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.
[ This comment has been converted into a solution ]
Log in to reply
The least value of K is obtained for a vector in the direction of the eigenvector belonging to the smallest eigenvalue
Why is this so?
Log in to reply
The transformed coordinates ( w , x , y , z ) express the original vector ( a , b , c , d ) as a linear combination of the eigenvectors.
We know that w 2 , x 2 , y 2 , z 2 add up to one, but the question is how we should distribute this total over the four coordinates. It matters because in the expression for K , each of them has a different weight.
Coordinate w corresponds to the eigenvector with eigenvalue − 3 , which is the smallest of the bunch. To minimize K , we want w 2 to be maximal, i.e. ( w , x , y , z ) = ( ± 1 , 0 , 0 , 0 ) .
Likewise, coordinate z corresponds to the greatest eigenvalue 2 + 5 . Maximizing K requires maximizing z 2 , i.e. ( w , x , y , z ) = ( 0 , 0 , 0 , ± 1 ) .
I did it basically the same way, but I used the symmetry of the problem to build in the following simplification:
for K m a x , a = b and c = d , and for K m i n , a = − b and c = − d .
Thus, I could use the substitutions ∣ a ∣ = x , ∣ c ∣ = y to create two simpler sets of coupled equations to solve: K m i n = − x 2 − 3 y 2 , x 2 + y 2 = 2 1 , and K m a x = x 2 + 4 x y + 3 y 2 , x 2 + y 2 = 2 1 .
I substituted y = 2 1 − x 2 , took the derivatives of the resulting expressions for K m i n / m a x , and found the roots of the resulting equations in both cases, which led to the same answers given in the OP.
Can it be done without including the variable x ?
Log in to reply
Including the variable x makes the solution more "natural". The other approach of finding K m a x is: Let t = c 2 + d 2 then K ≤ f ( t ) = t + 2 t ( 1 − t ) , t ∈ [ 0 , 1 ]
Can you explain when you found k min how did you get to know a=b and c= -d.
Also in final Maxima expression including x how did you make x+2= 2x^2 +3x?
Please provide the reasoning.
Log in to reply
Someone please clarify .
K min is achieved when inequalities at line 3 and line 7 (from the top) become equalies, or a + b + c + d = a = b = c + d = 0 . In finding K max, we choose x such that coefficients of ( a 2 + b 2 ) and ( c 2 + d 2 ) are the same, hence take advantage of ( a 2 + b 2 + c 2 + d 2 = 1 ) . You can solve with x = 2 5 − 1 from the beginning, but it doesn't look natural.
Here is my alternative approach, using linear algebra:
K = 2 1 ( a b c d ) ⎝ ⎜ ⎜ ⎛ 0 1 1 1 1 0 1 1 1 1 0 3 1 1 3 0 ⎠ ⎟ ⎟ ⎞ ⎝ ⎜ ⎜ ⎛ a b c d ⎠ ⎟ ⎟ ⎞ .
We determine the eigenvalues of the matrix by solving ∣ ∣ ∣ ∣ ∣ ∣ ∣ ∣ − λ 1 1 1 1 − λ 1 1 1 1 − λ 3 1 1 3 − λ ∣ ∣ ∣ ∣ ∣ ∣ ∣ ∣ = 0 ; λ 4 − 1 4 λ 2 − 1 6 λ − 3 = 0 ; ( λ + 1 ) ( λ 3 − λ 2 − 1 3 λ − 3 ) = 0 ; ( λ + 1 ) ( λ + 3 ) ( λ 2 − 4 λ − 1 ) = 0 ; ( λ + 1 ) ( λ + 3 ) ( λ − 2 + 5 ) ( λ − 2 − 5 ) = 0 ; λ = − 3 , − 1 , 2 − 5 , 2 + 5 . Thus we can rewrite K using an eigenvalue basis as follows: K = 2 1 ( ( − 3 ) w 2 + ( − 1 ) x 2 + ( 2 − 5 ) y 2 + ( 2 + 5 ) z 2 ) , where ( w , … z ) are the ( a , … , d ) expressed in a basis of orthonormal eigenvectors. We have w 2 + x 2 + y 2 + z 2 = 1 .
The least value of K is obtained for a vector in the direction of the eigenvector belonging to the smallest eigenvalue, etc. Thus: m = 2 1 ⋅ ( − 3 ) ⋅ 1 2 = − 1 . 5 ; M = 2 1 ⋅ ( 2 + 5 ) ⋅ 1 2 = 2 . 1 1 8 ; M + m = 2 5 − 1 = − 1 . 5 + 2 . 1 1 8 = 0 . 6 1 8 .
@Arjen Vreugdenhil , we really liked your comment, and have converted it into a solution.
In this reasoning it is essential that the matrix is symmetric . A symmetric matrix M can be uniquely factored in the form M = U T D U , where U is orthonormal and D is diagonal if the eigenvalues are real.
Thus the form K = a T M a becomes K = a T U T D U a = ( U a ) T D ( U a ) = w T D w . The equation a 2 + b 2 + c 2 + d 2 = 1 means a T a = 1 . Therefore w T w = ( U − 1 a ) T U − 1 a = a T U U − 1 a = a T a = 1 , where we used ( U − 1 ) T = U because U is orthonormal.
The transformation is w = 2 1 b + 2 5 5 − 5 c − 2 5 5 + 5 d ; x = − 2 1 b + 2 5 5 − 5 c − 2 5 5 + 5 d ; y = − 2 1 a + 2 5 5 + 5 c + 2 5 5 − 5 d ; z = 2 1 a + 2 5 5 + 5 c + 2 5 5 − 5 d . Minimum: ⎝ ⎜ ⎜ ⎛ a b c d ⎠ ⎟ ⎟ ⎞ = 2 1 ⎝ ⎜ ⎜ ⎛ 0 0 − 1 1 ⎠ ⎟ ⎟ ⎞ ; maximum: ⎝ ⎜ ⎜ ⎛ a b c d ⎠ ⎟ ⎟ ⎞ = 2 5 1 ⎝ ⎜ ⎜ ⎜ ⎛ 5 + 5 5 + 5 5 − 5 5 − 5 ⎠ ⎟ ⎟ ⎟ ⎞ .
Problem Loading...
Note Loading...
Set Loading...
Notice that
( a + b + c + d ) 2 ≥ 0 ⟹ 1 + 2 ( a b + a c + a d + b c + b d + c d ) ≥ 0 ⟹ a b + a c + a d + b c + b d + c d ≥ − 2 1 ⟹ K ≥ − 2 1 + 2 c d .
Also, ( c + d ) 2 + a 2 + b 2 ≥ 0 ⟹ 2 c d ≥ − ( a 2 + b 2 + c 2 + d 2 ) = − 1
Hence, K ≥ − 2 1 + 2 c d ≥ − 2 3 .
K = − 2 3 when a = b = 0 , c = − d , c 2 + d 2 = 1 .
So K m i n = − 2 3 when ( a , b , c , d ) = ( 0 , 0 , ± 2 1 , ∓ 2 1 ) .
By applying the popular inequality 2 a b ≤ a 2 + b 2 , with some real positive x we will have
2 x K = ( 2 a b ) x + 2 a ( x c ) + 2 a ( x d ) + 2 b ( x c ) + 2 b ( x d ) + ( 2 c d ) ( 3 x ) ≤ ( a 2 + b 2 ) x + a 2 + ( x c ) 2 + a 2 + ( x d ) 2 + b 2 + ( x c ) 2 + b 2 + ( x d ) 2 + ( c 2 + d 2 ) ( 3 x ) ⟹ 2 x K ≤ ( x + 2 ) ( a 2 + b 2 ) + ( 2 x 2 + 3 x ) ( c 2 + d 2 )
Now let choose x > 0 such that x + 2 = 2 x 2 + 3 x ⟺ x 2 + x − 1 = 0 ⟹ x = 2 5 − 1
Then
2 x K ≤ ( x + 2 ) ( a 2 + b 2 + c 2 + d 2 ) = x + 2 ⟹ K ≤ 2 x x + 2 = 2 5 + 2 .
K = 2 5 + 2 when
⎩ ⎪ ⎨ ⎪ ⎧ a = b = x c = x d x = 2 5 − 1 a 2 + b 2 + c 2 + d 2 = 1 ⟺ { a 2 = b 2 = 2 0 5 − 5 c 2 = d 2 = 2 0 5 + 5
So K m a x = 2 2 + 5 when ⎩ ⎨ ⎧ a = b = ± 2 0 5 − 5 c = d = ± 2 0 5 + 5 .
Finally, ( m + M ) = 2 2 + 5 − 3 = 2 5 − 1 .