Abstract
Let \(\left( X,\rho \right)\) be a complete matrix space, the nonlinear mapping \(\varphi:I\subset X\rightarrow X\) and the equation \(x=\varphi \left(x\right) \) with solution \(x^{\ast}\). We consider another application, \(F:X^{k}\rightarrow X\) for which we assume the diagonal coincides with \(\varphi\): \(F(x,…,x)=\varphi(x)\). In order to solve the mentioned equation we consider the iterative method \[x_{n+1}=F\left(x_{n},x_{n-1},\cdots,x_{n-k+1}\right),\] \(n=k-1,k,…\) Let \(i_{0},i_{1},….,i_{k-1}\) be a permutation of the numbers \(0,1,…,k-1\) and therefore \(i_{0}-n-k-1,~i_{1}+n-k+1,…,i_{k-1}+n-k+1\) a permutation of the numbers \(n-k+1,\ n-k+2,…,n\). Among the class of methods given by \[x_{n+1}=F\left( x_{i_{0}+n-k+1},x_{i_{1}+n-k+1},…,x_{i_{k-1}+n-k+1}\right) \] we determine the method for which the difference \(\mathcal{\rho}\left( x^{\ast},x_{n+1}\right)\) is the smallest.
Authors
Ion Păvăloiu
(Tiberiu Popoviciu Institute of Numerical Analysis)
Ioan Şerb
(Tiberiu Popoviciu Institute of Numerical Analysis)
Title
Original title (in French)
Sur des méthodes itératives optimales
English translation of the title
On optimal iterative methods
Keywords
multistep successive approximations; optimal iterative methods
Cite this paper as:
I. Păvăloiu, I. Şerb, Sur des méthodes itératives optimales, Seminar on functional analysis and numerical methods, Preprint no. 1 (1983), pp. 175-182 (in French).
About this paper
Journal
Preprint
Publisher Name
Faculty of Mathematics and Physics
Research Seminars
DOI
Not available yet.
References
[1] I. Pavaloiu, Introducere in teoria aproximarii solutiilor ecuatiilor, Ed. Dacia, Cluj-Napoca, 1976.
[2] I. Pavaloiu, Rezolvarea ecuatiilor prin interpolare, Ed. Dacia, Cluj-Napoca, 1981.
[3] Weinischke, J. H., Uber eine Klasse von Iterationsverfahren, Num. Math., 6 (1964), 395–404.
Paper (preprint) in HTML form
"Babeş-Bolyai" University
Faculty of Mathematics and Physics
Research Seminars
Seminar on Functional Analysis and Numerical Methods
Preprint Nr.1, 1983, pp. 175-182
On Optimal Iterative Methods
In this work we study iterative methods for solving the equation:
| (1) |
Oris a given application andis a complete metric space.
A l’application we will attach an application, which has the following property:
| (2) |
To solve equation ( 1 ) we consider the multi-step iterative method, of the form:
| (3) |
Andis a permutation of numbers, SOwill be a permutation that matches the numbers.
For such a permutation we obtain a new multi-step iterative method:
| (4) |
Or
In this way, from the given application, we obtain iterative methods, which are generally distinct.
If under certain conditions imposed on the application, all these iterative methods converge to a solution of equation ( 1 ), the problem arises of choosing the method for which the best error bound is obtained.
In the following we will assume thatverifies the following relationship:
| (5) |
whatever the elements, OrAndare given and
We note that ifverifies conditions ( 2 ) and ( 5 ) thenchecks the condition:
and the fact that, it follows thatis a contraction, that is, equation ( 1 ) has only one solution.
On the other hand, if we designate byall functionsverifying conditions ( 2 ) and ( 5 ), then ifit results.
Forwe deduce from Hölder's inequality:
SO.
Let's consider the numbers, , which verify the relations:
| (6) |
And
| (6′) |
We now consider the equations
| (7) |
| (8) |
And
| (9) |
Oris any permutation of the numbers .
LEMME.
Demonstration.
Eithera permutation of numbersand eitherthe biggest clue for which.
We obviously haveAnd.
Consider the function. SOAnd .
SOhas a positive root in the interval . It follows thathas a root in the interval.
The uniqueness of this root follows immediately by considering the equation, for whichandAnd.
To prove inequality ( 10 ) it suffices to show that And. In fact, we have:
becauseand from ( 6 ) it follows that
for each.
Similarly, we show that
Now we sort the coefficientsfrom ( 5 ) in descending order, that is:
| (11) |
and we write
| (11′) |
We consider the application, given by the relation
| (12) |
Oris the permutation of numbers which corresponds to the relation ( 11 ). We then obtain from ( 5 ), for, the condition
| (13) |
whatever the elementsOr
To solve equation ( 1 ) we consider the following iterative method:
| (14) |
Or.
If we write, then from ( 13 ) and ( 14 ) we obtain
| (15) |
If in equation ( 7 ) are given by ( 11 ′ ) and ifis the positive root of equation ( 7 ), then.
Eithera positive constant such that:
From ( 15 ) we obtain:
. Sinceit follows that the followingis fundamental. Let us designate byits limit. It is obvious that is the unique solution of equation ( 1 ) and
| (16) |
∎
So, using the lemma we obtain the following theorem:
THEOREM .
Let's suppose thatand that we give theiterative methods of the form ( 4 ), attached to.
Remarks .
1) Because, from theit follows thatis a contraction, with the Lipschitz constant, it is clear that the iterative method
| (17) |
converges to the solutionfrom equation ( 1 ). Furthermore, we have
| (18) |
We will demonstrate that, where a is the solution to equation ( 7 ), with given by the formula ( 11 ′ ).
On a:
from which it follows that.
Be it nowAnd. Formulas ( 16 ) and ( 18 ) valid for any metric and for each pairwith the above properties, show us that for the iterative method ( 17 ) we obtain a better error delimitation than in the case of other methods of the form ( 4 ). This does not mean that for each choice of the initial points and each applicationmethod ( 17 ) converges faster than all iterative methods of the form ( 4 ).
For example,, . On a
for each; And.
Example .
Either
Either
| (22) |
| (23) |
| (24) |
In this case we obtain the formulas:
| (22′) |
| (23′) |
where a is the positive root of the equation, that's to say. In the same way we obtain
| (24′) |
oris the positive root of the characteristic equation that's to say. It is clear that
Bibliography
-
[1]
I. Pavaloiu,
††margin:
clickable
clickable Introduction to the theory of approximation of solutions of equations , Ed. Dacia, Cluj-Napoca, 1976. - [2] I. Păvăloiu, Solving equations by interpolation , Dacia Publishing House, Cluj-Napoca, 1981.
- [3] Weinischke, JH, On a class of iteration methods , Num. Math., 6 (1964), 395–404.
