Newton's method of estimation - using derivatives

Click For Summary
SUMMARY

Newton's method for estimating real roots of equations is derived from the tangent line approximation. The process begins with an initial guess x1, and iteratively refines this guess using the formula xn+1 = xn - f(xn)/f’(xn). This method requires that the function f is differentiable and that its derivative f’(xn) is non-zero. The discussion also illustrates how to apply this method to approximate the square root of a positive number k using the function f(x) = x^2 - k.

PREREQUISITES
  • Understanding of calculus, specifically derivatives and tangent lines.
  • Familiarity with Newton's method and its iterative process.
  • Knowledge of differentiable functions and their properties.
  • Basic algebra for manipulating equations and solving for variables.
NEXT STEPS
  • Study the convergence criteria for Newton's method to ensure accurate approximations.
  • Explore the application of Newton's method in solving non-linear equations.
  • Learn about the limitations and potential pitfalls of using Newton's method.
  • Investigate alternative root-finding algorithms such as the Bisection method and Secant method.
USEFUL FOR

Students in calculus, mathematicians, engineers, and anyone interested in numerical methods for solving equations.

Bruce3
Messages
2
Reaction score
0

Homework Statement



Newton devised the following method for approximating a real root of the equation f(x) = 0. i.e. a real number for which f(r) = 0. We begin by guessing an approximation, say x1, to the real root r.

(i) Find the equation of the line tangent to the graph of y = f(x) at the point (x1,f(x1)). Assume f is differentiable and f’(x1) ≠ 0.

(ii) Newton felt that the x-intercept of the tangent line in part(i) would be a better approximation to r than x1. Label the point of intersection of the tangent line in (i) with the x-axis (x2,0). Find x2 in terms of x1, f(x1), and f’(x1).

(iii) The above procedure may now be repeated using the tangent line at (x2,f(x2)). If f’(x2) ≠ 0 this leads to a third approximation, x3, where (x3,0) is the point of intersection of the x-axis with the tangent line at (x2,f(x2)). Find x3 in terms of x2, f(x2), and f’(x2).

If done correctly, you have just derived Newton’s method, which can be written as:
To approximate the real root of r of f(x)=0, begin by guessing an initial approximation to r, say x1, For n = 1, 2, 3, … let xn+1 = xn – f(xn)/f’(xn). Under “good” conditions the sequence xn will converge to r.

(iv) Approximate k^.5 for k > 0 by applying Newton’s method to f(x) = x^2 – k with the initial guess x1, You have just derived the method Heron used around 100 A.D. to approximate k^.5.

Homework Equations


To approximate the real root of r of f(x)=0, begin by guessing an initial approximation to r, say x1, For n = 1, 2, 3, … let xn+1 = xn – f(xn)/f’(xn). Under “good” conditions the sequence xn will converge to r.

My work
(i) for (x1, f(x1))
y – f(x1) = f’(x1)(x-x1)
y = f’(x1)(x-x1)+f(x1)
(ii) for (x2,0)
0=f’(x1)(x2-x1) + f(x1)
-f(x1) = f’(x1)(x2-x1)
x2 – x1 = -f(x1)/f’(x1)
x2 = x1 – f(x1)/f’(x1)
(iii) for (x2, f(x2))
y – f(x2) = f’(x2)(x-x2)
y = f’(x2)(x-x2)+f(x2)
(ii) for (x3,0)
0=f’(x2)(x3-x2) + f(x2)
-f(x2) = f’(x2)(x3-x2)
x3 – x2 = -f(x2)/f’(x2)
x3 = x2 – f(x2)/f’(x2)
(iv)
f(x) = x^2 –k
f’(x) = 2x
xn+1 = xn – (n^2-k)/(2n)
xn+1 = xn – n/2 – k/(2x)
xn+1 – xn + n/2 = – k/(2x)
(xn+1 – xn + n/2)*(-2n) = k

Any help on the work is greatly appreciated.
 
Last edited by a moderator:
Physics news on Phys.org
Any help on the work is greatly appreciated.
Cool! Um... what seems to be the problem?
...
f(x) = x^2 –k
f’(x) = 2x
xn+1 = xn – (n^2-k)/(2n)
... this is where you lose me: what is your reasoning here?
 

Similar threads

  • · Replies 26 ·
Replies
26
Views
3K
  • · Replies 1 ·
Replies
1
Views
2K
Replies
2
Views
1K
  • · Replies 9 ·
Replies
9
Views
4K
  • · Replies 6 ·
Replies
6
Views
2K
Replies
14
Views
2K
  • · Replies 5 ·
Replies
5
Views
4K
Replies
9
Views
3K
  • · Replies 1 ·
Replies
1
Views
2K
  • · Replies 11 ·
Replies
11
Views
3K