Trouble with taylor's Inequality

  • Thread starter Thread starter Winzer
  • Start date Start date
  • Tags Tags
    Inequality
Click For Summary

Homework Help Overview

The discussion revolves around Taylor's Inequality, specifically the computation of the remainder term |R_{n}| in Taylor series expansions. Participants are exploring how to determine the upper bound M for the (n+1)th derivative of a function within a specified interval.

Discussion Character

  • Exploratory, Conceptual clarification, Mathematical reasoning

Approaches and Questions Raised

  • Participants discuss methods for finding the maximum value M of the (n+1)th derivative, both through calculus and without it. There are attempts to apply Taylor's theorem to specific functions, such as f(x) = √x and f(x) = √(x+1), while questioning the validity of their upper bounds and the implications of their calculations.

Discussion Status

The discussion is active, with participants sharing insights on how to compute the remainder term and upper bounds for different functions. Some have provided specific examples and calculations, while others are seeking clarification on the concepts and methods involved.

Contextual Notes

Participants are working within the constraints of homework guidelines, focusing on understanding the theoretical aspects of Taylor's Inequality and the implications of their calculations without arriving at definitive conclusions.

Winzer
Messages
597
Reaction score
0

Homework Statement


Ok so I have been having trouble computing [tex]|R_{n}|[/tex]
in problems.

Homework Equations


[tex]|R_{n}|\leq \frac{M}{(n+1)!}|x-a|[/tex]^(n+1)
Edit:
[tex]M\geq f[/tex]^(n+1) (h)

The Attempt at a Solution


Can some one walk me through calculatng some generic solution.
Like finding M and all that? I need more insight on what it actually does too.
Thanks
 
Last edited:
Physics news on Phys.org
if f is differentiable, you can use the methods of calculus to find the extremas of f and hence its absolute max. such a number is as M in your post.

in other cases, you can find such an M w/o calculus. If foor instance, f(x)=4sin(x). Well, |sin(x)|[itex]\leq[/itex]1, so |f(x)|[itex]\leq[/itex]4. So M=4 will do
 
If you use the nth Taylor polynomial for ex, about x= 0, the error is less than
[tex]\frac{M}{(n+1)!}|x|^{n+1}[/tex]
where M is an upperbound on the n+1 derivative between 0 and x. Of course, every derivative of ex is ex itself so we just need an upperbound on that. For example if x= 1, We know the derivative is less than e so we can use that as an upper bound. The error is less than
[tex]\frac{e}{(n+1)!}[/itex]<br /> <br /> Calculating errors for the Taylor Polynomials for sine and cosine is even easier: all derivatives are either sine or cosine and 1 is always an upper bound.[/tex]
 
Ok so suppose I have my function [tex]f(x)=\sqrt(x)[/tex] and the interval [.5,1.5] using n=2 centered at 1. So:
[tex]|R_{2}|\leq \frac{M}{3!} |x-1|^3[/tex]
[tex]f^3(x)= \frac{3}{8x^(5/2)}\leq M[/tex]. My upperbound on [tex]f^3(x)[/tex] is 1.5 so [tex]f^3(1.5)=\frac{\sqrt(3)}{72}[/tex]
So: [tex]|R_{2}|\leq \frac{\sqrt(3)}{72*3!} |x-1|^3[/tex] X=1.5
So: So: [tex]|R_{2}|\leq \frac{\sqrt(3)}{72*3!} |.5|^3[/tex]
How does that sound?
 
Last edited:
Say [tex]f(x) = \sqrt{x+1}[/tex] centered at 0, the reason why I am doing it this way is because you can get same result upon a basic substitution. On [tex](-1,1)[/tex].

Thus,
[tex]f(x) = (x+1)^{1/2} \implies f'(0)=1[/tex]
[tex]f'(x) = \frac{1}{2}(x+1)^{-1/2} \implies f'(0)=\frac{1}{2}[/tex]
[tex]f''(x) = - \frac{1}{4}(x+1)^{-3/2} \implies f''(0) = -\frac{1}{4}[/tex]

That means for some [tex]y[/tex] between [tex]x\not =0[/tex] and [tex]0[/tex] we have:
[tex]R_2(x) = f(x) - T_2(x) = \frac{f'''(y)}{3!}x^3[/tex]

Now,
[tex]\left| \frac{f'''(y)}{3!}x^3 \right| = \left| \frac{3}{3!\cdot 8}\cdot \frac{(y+1)^{-5/2}}{1} \cdot x^3 \right| \leq \frac{3}{3!\cdot 8}[/tex]
 

Similar threads

Replies
12
Views
2K
  • · Replies 7 ·
Replies
7
Views
3K
Replies
2
Views
2K
  • · Replies 8 ·
Replies
8
Views
2K
  • · Replies 6 ·
Replies
6
Views
2K
Replies
12
Views
2K
Replies
1
Views
1K
  • · Replies 3 ·
Replies
3
Views
2K
  • · Replies 27 ·
Replies
27
Views
4K
Replies
5
Views
2K