Find the smallest constant C such that for every polynomial P ( x ) of degree 3 that has a root in the interval [ 0 , 1 ] ,
∫ 0 1 ∣ P ( x ) ∣ d x ≤ C ⋅ { x ∈ [ 0 , 1 ] } max ∣ P ( x ) ∣ .
Enter ⌊ 1 0 0 0 C ⌋ .
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.
Beautiful proof !! But I think there is a little (but important) error (I might be wrong...) It is not K ≥ 5 / 6 but K ≤ 5 / 6 and same for C at the end.
I had the value 5 / 6 (numerical tests, iterating process that maximises the integral divided by the max by changing the coefficients of P in the direction indicated by the gradient) and the idea that P ( 0 ) has to be 0 while the maximum has to be reached at 1 and in the "middle" to maximised the time near the maximum during the wavy part of the cubic but I wasn't able to prove it. Your formula that gives ∫ 0 1 P ( x ) d x in terms of P ( 0 ) , P ( 1 / 2 ) , P ( 1 ) is incredibly efficient for this problem !
Log in to reply
After a further look, I agree, the ≥ symbol was a typo, and should have been ≤ . The polynomial Q 0 is what gives K ≥ 6 5 , and hence K = 6 5 .
Problem Loading...
Note Loading...
Set Loading...
Let K be the smallest positive constant such that ∫ 0 1 P ( x ) d x ≤ K m a x x ∈ [ 0 , 1 ] P ( x ) for all cubic polynomials P such that P ( 0 ) = 0 and 0 ≤ P ( x ) ≤ 1 for all 0 ≤ x ≤ 1 .
If P ( x ) = a x 3 + b x 2 + c x + d is a cubic polynomial,then ∫ 0 1 P ( x ) d x = 4 1 a + 3 1 b + 2 1 c + d = 6 1 P ( 0 ) + 3 2 P ( 2 1 ) + 6 1 P ( 1 ) Thus, if P ( x ) is a cubic polynomial such that P ( 0 ) = 0 and 0 ≤ P ( x ) ≤ 1 for all 0 ≤ x ≤ 1 , then ∫ 0 1 ∣ P ( x ) ∣ d x = 3 2 P ( 2 1 ) + 6 1 P ( 1 ) ≤ 3 2 + 6 1 = 6 5 and hence K ≤ 6 5 . However, if we consider the polynomial Q 0 ( x ) = 1 − 4 ( x − 1 ) ( x − 2 1 ) 2 then Q 0 ( x ) is cubic with Q 0 ( 0 ) = 0 and Q 0 ( 1 ) = Q 0 ( 2 1 ) = 1 , and Q 0 has a turning point at x = 2 1 . The second turning point of Q 0 occurs at x = 6 5 , and Q 0 ( 6 5 ) = 2 7 2 5 > 0 . Thus we deduce that Q 0 ( x ) is a cubic polynomial with Q 0 ( 0 ) = 0 and 0 ≤ Q 0 ( x ) ≤ 1 for all 0 ≤ x ≤ 1 , and moreover that ∫ 0 1 ∣ Q 0 ( x ) ∣ d x = 6 5 . Thus we deduce that K = 6 5 .
If we now assume that the cubic Q ( x ) is single-signed throughout [ 0 , 1 ] and that either Q ( 0 ) = 0 or Q ( 1 ) = 0 , then a single linear change of variable, and scaling Q suitably, shows that ∫ 0 1 ∣ Q ( x ) ∣ d x ≤ 6 5 x ∈ [ 0 , 1 ] max ∣ Q ( x ) ∣
If [ u , v ] is a subinterval of [ 0 , 1 ] , and if the cubic Q ( x ) is single-signed on [ u , v ] , and vanishes at one of u or v , then a simple linear change of variable shows us that ∫ u v ∣ Q ( x ) ∣ d x ≤ 6 5 ( v − u ) m a x x ∈ [ u , v ] ∣ Q ( x ) ∣ ≤ 6 5 ( v − u ) m a x x ∈ [ 0 , 1 ] ∣ Q ( x ) ∣
Suppose now that P ( x ) is a cubic polynomial which has at least one root in [ 0 , 1 ] . We can then subdivide the interval [ 0 , 1 ] at the zeros of P ( x ) , and we therefore deduce, considering the integral of ∣ P ( x ) ∣ on each of these intervals separately, that ∫ 0 1 ∣ P ( x ) ∣ d x ≤ 6 5 m a x x ∈ [ 0 , 1 ] ∣ P ( x ) ∣ which tells us that C ≥ 6 5 . Considering Q 0 ( X ) again tells us that C = 6 5 . This makes the answer ⌊ 1 0 0 0 × 6 5 ⌋ = 8 3 3 .