Optimizing Polynomial Approximations for C2 Functions on Closed Intervals

  • Thread starter Thread starter Demon117
  • Start date Start date
  • Tags Tags
    Polynomial
Join the discussion
Ask a follow-up here, or get your own question answered by working scientists, mathematicians and engineers — people, not an autocomplete.
Real named experts · corrections over time · the nuance an AI answer skips
1 reply · 1K views
Demon117
Messages
162
Reaction score
1
1. Suppose that f:R-->R is of class C2. Given b>0 and a positive number [tex]\epsilon[/tex], show that there is a polynomial p such that

|p(x)-f(x)|<[tex]\epsilon[/tex], |p'(x)-f'(x)|<[tex]\epsilon[/tex], |p"(x)-f"(x)|<[tex]\epsilon[/tex] for all x in [0,b].



The Attempt at a Solution



First I choose a polynomial q that approximates f''. If |q - f''|<[tex]\eta[/tex] throughout [0,b], and if p is the polynomial such that p''=q, p(0)=f(0), and p'(0)=f'(0), then I come to this question: How big can |p' - f'| and |p - f| be in terms of [tex]\eta[/tex] and b? I think I am thinking about this correctly, but I cannot come to a conclusion. Should I use the definition of the derivative namely:

For all [tex]\epsilon[/tex]>0, there is a [tex]\delta[/tex]>0, such that when 0<|t-x|<[tex]\delta[/tex], this guarantees that |(p(t)-p(x))/(t-x) - L|<[tex]\epsilon[/tex]; where L is the derivative at x.
 
Physics news on Phys.org
Observe that [tex](p' - f')' = p'' - f''[/tex]; what does this tell you about the accumulated error in approximating [tex]f'[/tex] by [tex]p'[/tex], as you move across the interval from [tex]0[/tex] to [tex]b[/tex]?