Sur l’approximation des fonctions convexes d’ordre supérieur (I)

Abstrait

Traduction en anglais du titre

On the approximation of higher order convex functions (I)

Auteur(s)

Mots-clés

PDF

Pour citer ce travail

T. Popoviciu, Sur l’approximation des fonctions convexes d’ordre supérieur (I), Mathematica, 10 (1935), pp. 49-54 (in French).

Sur ce travail

Journal

Mathematica

Publié par

Publié par les Éditions de l’Académie roumaine

DOI

Non disponible.

Print ISSN

Non disponible.

Online ISSN

Non disponible.

??

HTML forme du travail (preprint)

1935 e -Popoviciu- Mathematica - Sur l_approximation des fonctions convexes d_ordre superieur (I).pd

SUR L'APPROXIMATION DES FONCTIONS CONVEXES D'ORDRE SUPÉRIEUR.

ParTiberiu PopoviciuAncien t́lève de l'École Normale Supérieure

Reçue le 20 Mars 1934.
Soit f ( x ) f ( x ) f(x)f(x)f(x) une fonction définie dans l'intervalle ( 0 , 1 ) ( 0 , 1 ) (0,1)(0,1)(0,1).
Définissons l'expression [ x 1 , x 2 , , x n + 1 ; f ] x 1 , x 2 , , x n + 1 ; f [x_(1),x_(2),dots,x_(n+1);f]\left[x_{1}, x_{2}, \ldots, x_{n+1} ; f\right][x1,x2,,xn+1;f] par la relation de récurrence
[ x 1 , x 2 , , x n + 1 ; f ] = [ x 2 , x 3 , , x n + 1 ; f ] [ x 1 , x 2 , , x n ; f ] x n + 1 x 1 [ x 1 ; f ] = f ( x 1 ) . x 1 , x 2 , , x n + 1 ; f = x 2 , x 3 , , x n + 1 ; f x 1 , x 2 , , x n ; f x n + 1 x 1 x 1 ; f = f x 1 . {:[[x_(1),x_(2),dots,x_(n+1);f]=([x_(2),x_(3),dots,x_(n+1);f]-[x_(1),x_(2),dots,x_(n);f])/(x_(n+1)-x_(1))],[[x_(1);f]=f(x_(1)).]:}\begin{gathered} {\left[x_{1}, x_{2}, \ldots, x_{n+1} ; f\right]=\frac{\left[x_{2}, x_{3}, \ldots, x_{n+1} ; f\right]-\left[x_{1}, x_{2}, \ldots, x_{n} ; f\right]}{x_{n+1}-x_{1}}} \\ {\left[x_{1} ; f\right]=f\left(x_{1}\right) .} \end{gathered}[x1,x2,,xn+1;f]=[x2,x3,,xn+1;f][x1,x2,,xn;f]xn+1x1[x1;f]=f(x1).
Le quotient [ x 1 , x 2 , x n + 1 ; f ] x 1 , x 2 , x n + 1 ; f [x_(1),x_(2),dotsx_(n+1);f]\left[x_{1}, x_{2}, \ldots x_{n+1} ; f\right][x1,x2,xn+1;f] est la différence divisée d'ordre n n nnn de la fonction f ( x ) f ( x ) f(x)f(x)f(x) par rapport aux points distincts x 1 , x 2 , x n + 1 x 1 , x 2 , x n + 1 x_(1),x_(2),dotsx_(n+1)x_{1}, x_{2}, \ldots x_{n+1}x1,x2,xn+1. Soit
lim [ dans ( 0 , 1 ) ] x 1 , x 2 , , x n + 1 ; f ] ∣= Δ n [ f ] . lim ¯ [ dans ( 0 , 1 ) ] x 1 , x 2 , , x n + 1 ; f ∣= Δ n [ f ] . {: bar(lim)_([dans(0,1)])∣x_(1),x_(2),dots,x_(n+1);f]∣=Delta_(n)[f].\left.\varlimsup_{[\operatorname{dans}(0,1)]} \mid x_{1}, x_{2}, \ldots, x_{n+1} ; f\right] \mid=\Delta_{n}[f] .lim[dans(0,1)]x1,x2,,xn+1;f]∣=Δn[f].
Δ n [ f ] Δ n [ f ] Delta_(n)[f]\Delta_{n}[f]Δn[f] est la nème borne de f ( x ) f ( x ) f(x)f(x)f(x) dans ( 0 , 1 ) ( 0 , 1 ) (0,1)(0,1)(0,1).
Considérons les points
(1)
x 1 < x 2 < . < x m x 1 < x 2 < . < x m x_(1) < x_(2) < dots. < x_(m)x_{1}<x_{2}<\ldots .<x_{m}x1<x2<.<xm
alors
v m = i = 1 m n 1 | [ x i + 1 , x i + 2 , , x i + n + 1 ; f ] [ x i , x i + 1 , , x i + n ; f ] | v m = i = 1 m n 1 x i + 1 , x i + 2 , , x i + n + 1 ; f x i , x i + 1 , , x i + n ; f v_(m)=sum_(i=1)^(m-n-1)|[x_(i+1),x_(i+2),dots,x_(i+n+1);f]-[x_(i),x_(i+1),dots,x_(i+n);f]|v_{m}=\sum_{i=1}^{m-n-1}\left|\left[x_{i+1}, x_{i+2}, \ldots, x_{i+n+1} ; f\right]-\left[x_{i}, x_{i+1}, \ldots, x_{i+n} ; f\right]\right|vm=i=1mn1|[xi+1,xi+2,,xi+n+1;f][xi,xi+1,,xi+n;f]|
est la nème variation de f ( x ) f ( x ) f(x)f(x)f(x) sur les points (1).
Si nous posons
lim [ dans ( 0 , 1 ) ] v m = n [ f ] lim ¯ [ dans  ( 0 , 1 ) ] v m = n [ f ] bar(lim)_(["dans "(0,1)])v_(m)=grad_(n)[f]\varlimsup_{[\text {dans }(0,1)]} v_{m}=\nabla_{n}[f]lim[dans (0,1)]vm=n[f]
alors V n [ f ] V n [ f ] V_(n)[f]V_{n}[f]Vn[f] est la nème variation totale de f ( x ) f ( x ) f(x)f(x)f(x) dans ( 0 , 1 ) ( 0 , 1 ) (0,1)(0,1)(0,1).
Nous dirons enfin que la fonction f ( x ) f ( x ) f(x)f(x)f(x) est convexe, non-concave, polynomiale, non-convexe ou concave d'ordre n n nnn dans ( 0 , 1 ) ( 0 , 1 ) (0,1)(0,1)(0,1) si ses différences divisées d'ordre n + 1 n + 1 n+1n+1n+1 sur tout groupe de n + 2 n + 2 n+2n+2n+2 points ( 0 , 1 ) ( 0 , 1 ) (0,1)(0,1)(0,1) sont > 0 , 0 , = 0 0 > 0 , 0 , = 0 0 > 0, >= 0,=0 <= 0>0, \geq 0,=0 \leq 0>0,0,=00 ou < 0 < 0 < 0<0<0.
Cies fonctions forment la classe des fonction d'ordre n n nnn.
Nous dirons que la fonction est de la classe ( a , b , c , a , b , c , a,b,c,dotsa, b, c, \ldotsa,b,c, ) si elle possède des propriétés d'ordre a , b , c , a , b , c , a,b,c,dotsa, b, c, \ldotsa,b,c, d'une nature de convexité déterminée. Pour l'uniformité on peut appeler fonction d'ordre - 1 1 1\mathbf{1}1 toute fonction ne changeant pas de signe ( 1 ).
  1. Le polynome
P n = P n ( x ; f ) = i = 0 n f ( i n ) ( n i ) x i ( 1 x ) n t P n = P n ( x ; f ) = i = 0 n f i n ( n i ) x i ( 1 x ) n t P_(n)=P_(n)(x;f)=sum_(i=0)^(n)f((i)/(n))*((n)/(i))x^(i)(1-x)^(n-t)\mathrm{P}_{n}=\mathrm{P}_{n}(x ; f)=\sum_{i=0}^{n} f\left(\frac{i}{n}\right) \cdot\binom{n}{i} x^{i}(1-x)^{n-t}Pn=Pn(x;f)=i=0nf(in)(ni)xi(1x)nt
est le polynome de M. S. Bernstein de degré n n nnn de la fonction donnée.
Posons pour abréger les notations
Δ k i [ i + k n , i + k 1 n , , i n ; f ] , i = 0 , 1 , , n k , k = 1 , 2 , Δ k i i + k n , i + k 1 n , , i n ; f , i = 0 , 1 , , n k , k = 1 , 2 , Delta_(k)^(i)◻[(i+k)/(n),(i+k-1)/(n),dots,(i)/(n);f],quad i=0,1,dots,n-k,k=1,2,dots\Delta_{k}^{i} \square\left[\frac{i+k}{n}, \frac{i+k-1}{n}, \ldots, \frac{i}{n} ; f\right], \quad i=0,1, \ldots, n-k, k=1,2, \ldotsΔki[i+kn,i+k1n,,in;f],i=0,1,,nk,k=1,2,
Un calcul simple nous donne
d P n d x = i = 0 n 1 Δ 1 i ( n 1 i ) x i ( 1 x i ) n 1 i d P n d x = i = 0 n 1 Δ 1 i ( n 1 i ) x i 1 x i n 1 i (dP_(n))/(dx)=sum_(i=0)^(n-1)Delta_(1)^(i)((n-1)/(i))x^(i)(1-x^(i))^(n-1-i)\frac{d \mathrm{P}_{n}}{d x}=\sum_{i=0}^{n-1} \Delta_{1}^{i}\binom{n-1}{i} x^{i}\left(1-x^{i}\right)^{n-1-i}dPndx=i=0n1Δ1i(n1i)xi(1xi)n1i
et en général
(2) d k P n d x k = k ! ( 1 1 n ) ( 1 2 n ) d k P n d x k = k ! 1 1 n 1 2 n (d^(k)P_(n))/(dx^(k))=k!(1-(1)/(n))(1-(2)/(n))dots\frac{d^{k} \mathrm{P}_{n}}{d x^{k}}=k!\left(1-\frac{1}{n}\right)\left(1-\frac{2}{n}\right) \ldotsdkPndxk=k!(11n)(12n)
( 1 k 1 n ) l = 0 n k Δ k i ( n k i ) x l ( 1 k ) n k 1 1 k 1 n l = 0 n k Δ k i ( n k i ) x l ( 1 k ) n k 1 dots(1-(k-1)/(n))sum_(l=0)^(n-k)Delta_(k)^(i)((n-k)/(i))x^(l)(1-k)^(n-k-1)\ldots\left(1-\frac{k-1}{n}\right) \sum_{l=0}^{n-k} \Delta_{k}^{i}\binom{n-k}{i} x^{l}(1-k)^{n-k-1}(1k1n)l=0nkΔki(nki)xl(1k)nk1
Nous avons
Δ 0 [ P n ] Δ 0 [ f ] Δ 0 P n Δ 0 [ f ] Delta_(0)[P_(n)] <= Delta_(0)[f]\Delta_{0}\left[\mathrm{P}_{n}\right] \leq \Delta_{0}[f]Δ0[Pn]Δ0[f]
Nous savons que ( 2 2 ^(2){ }^{2}2 )
k ! Δ k [ P n ] = Δ 0 [ P n ( k ) ] k ! Δ k P n = Δ 0 P n ( k ) k!Delta_(k)[P_(n)]=Delta_(0)[P_(n)^((k))]k!\Delta_{k}\left[\mathrm{P}_{n}\right]=\Delta_{0}\left[\mathrm{P}_{n}^{(k)}\right]k!Δk[Pn]=Δ0[Pn(k)]
donc
Δ ˙ k [ P n ] ( 1 1 n ) ( 1 2 n ) ( 1 k 1 n ) Δ k [ f ] Δ ˙ k P n 1 1 n 1 2 n 1 k 1 n Δ k [ f ] Delta^(˙)_(k)[P_(n)] <= (1-(1)/(n))(1-(2)/(n))cdots(1-(k-1)/(n))*Delta_(k)[f]\dot{\Delta}_{k}\left[\mathrm{P}_{n}\right] \leq\left(1-\frac{1}{n}\right)\left(1-\frac{2}{n}\right) \cdots\left(1-\frac{k-1}{n}\right) \cdot \Delta_{k}[f]Δ˙k[Pn](11n)(12n)(1k1n)Δk[f]
( 1 1 ^(1){ }^{1}1 ) Pour plus de détails voir notre Thèse Sur quelques propriétés des fonctions d'une ou de deux variables réelles. Paris 1933.
( 2 2 ^(2){ }^{2}2 ) Voir loc. cit. (1), p. 45.
On pcut encore écrire
Δ k [ P n ] Δ k [ f ] , k = 0 , 1 , Δ k [ P n ] < Δ k [ f ] , k > 1 , Δ k P n Δ k [ f ] , k = 0 , 1 , Δ k P n < Δ k [ f ] , k > 1 , Delta_(k)[P_(n)] <= Delta_(k)[f],quad k=0,1,quadDelta_(k)[P_(n)] < Delta_(k)[f],quad k > 1,\Delta_{k}\left[\mathrm{P}_{n}\right] \leq \Delta_{k}[f], \quad k=0,1, \quad \Delta_{k}\left[\mathrm{P}_{n}\right]<\Delta_{k}[f], \quad k>1,Δk[Pn]Δk[f],k=0,1,Δk[Pn]<Δk[f],k>1,
donc:
La kème borne du polynome P n ( x ; f ) P n ( x ; f ) P_(n)(x;f)P_{n}(x ; f)Pn(x;f) est au plus égale à celle de la fonction si k = 0 , 1 k = 0 , 1 k=0,1k=0,1k=0,1 et est plus petite si k < 1 k < 1 k < 1k<1k<1.
2. Nous pouvons écrire encore
d k P n d x k = k ! ( 1 1 n ) ( 1 2 n ) ( 1 k 1 n ) { Δ k n k + + ( n k ) i = 0 n k 1 [ Δ k i Δ k i + 1 ] ( n k 1 i ) x 1 t l ( 1 t ) n k i 1 d t d k P n d x k = k ! ( 1 1 n 1 2 n 1 k 1 n Δ k n k + + ( n k ) i = 0 n k 1 Δ k i Δ k i + 1 ( n k 1 i ) x 1 t l ( 1 t ) n k i 1 d t {:[(d^(k)P_(n))/(dx^(k))=k!(1-{:(1)/(n))(1-(2)/(n))cdots(1-(k-1)/(n))*{Delta_(k)^(n-k)+:}],[+(n-k)sum_(i=0)^(n-k-1)[Delta_(k)^(i)-Delta_(k)^(i+1)]((n-k-1)/(i))int_(x)^(1)t^(l)(1-t)^(n-k-i-1)dt]:}\begin{aligned} \frac{d^{k} \mathrm{P}_{n}}{d x^{k}}=k!(1- & \left.\frac{1}{n}\right)\left(1-\frac{2}{n}\right) \cdots\left(1-\frac{k-1}{n}\right) \cdot\left\{\Delta_{k}^{n-k}+\right. \\ & +(n-k) \sum_{i=0}^{n-k-1}\left[\Delta_{k}^{i}-\Delta_{k}^{i+1}\right]\binom{n-k-1}{i} \int_{x}^{1} t^{l}(1-t)^{n-k-i-1} d t \end{aligned}dkPndxk=k!(11n)(12n)(1k1n){Δknk++(nk)i=0nk1[ΔkiΔki+1](nk1i)x1tl(1t)nki1dt
et on en déduit
V k [ P n ] V k [ f ] , k = 0 , 1 , V k [ P n ] < V k [ f ] , k > 1 V k P n V k [ f ] , k = 0 , 1 , V k P n < V k [ f ] , k > 1 V_(k)[P_(n)] <= V_(k)[f],quad k=0,1,quadV_(k)[P_(n)] < V_(k)[f],quad k > 1\mathrm{V}_{k}\left[\mathrm{P}_{n}\right] \leq \mathrm{V}_{k}[f], \quad k=0,1, \quad \mathrm{~V}_{k}\left[\mathrm{P}_{n}\right]<\mathrm{V}_{k}[f], \quad k>1Vk[Pn]Vk[f],k=0,1, Vk[Pn]<Vk[f],k>1
donc:
La kème variation totale du polynome P n ( x ; f ) P n ( x ; f ) P_(n)(x;f)P_{n}(x ; f)Pn(x;f) est au plus égale d celle de la fonction si k = 0 , 1 k = 0 , 1 k=0,1k=0,1k=0,1 et est plus petitie si k > 1 k > 1 k > 1k>1k>1.
On peut exprimer les deux propriétés précédentes en disant que le polynome de M. S. Bernstein conserve les bornes et les variations.
3. La formule (2) nous montre encore que toutes les fois que f ( x ) f ( x ) f(x)f(x)f(x) jouit d'une propriété de convexité déterminée (d'ordre < n < n < n<n<n ) le polynome P n ( x ; f ) P n ( x ; f ) P_(n)(x;f)\mathrm{P}_{n}(x ; f)Pn(x;f) aura la même propriété. Nous supposons ici que la convexité et la polynomialité sont des cas particuliers de la non-concavité.
Nous exprimons cette propriété en disant que le polynome ed M. Bernstein conserve la classe de la fonction.
Une propriété démontrée dans notre travail cité ( 3 3 ^(3){ }^{3}3 ) permet d'établir que
Il existe des polynomes d'une classe donnée dans ( 0 , 1 ) ( 0 , 1 ) (0,1)(0,1)(0,1) cette classe contenant un nombre fini de conditions de convexité ou concavité choisies arbitrairement.
On sait enfin que si la fonction f ( x ) f ( x ) f(x)f(x)f(x) est continue dans ( 0 , 1 ) ( 0 , 1 ) (0,1)(0,1)(0,1) le polynome P n ( x ; f ) P n ( x ; f ) P_(n)(x;f)P_{n}(x ; f)Pn(x;f) tend uniformément vers f ( x ) f ( x ) f(x)f(x)f(x) dans tout l'intervalle pour n n n rarr oon \rightarrow \inftyn ( 4 4 ^(4){ }^{4}4 ). Nous avons donc la propriété suivante:
( 3 3 ^(3){ }^{3}3 ) Voir loc. cit. (1) p. 20.
(4) S. Bernstein Communications de la Soc. Math. de Karkow ser. 2, t. 13 (1912) (Citation du livre de MM. Pólya et Szegö „Aufgaben und Lehrsätze" t. I. p. 230). Voir aussi S. Wigert Arkiv för Mat. Astr. och, Physik t. 20, (1927), p, 1.
Toute fonction continue dans ( 0 , 1 ) ( 0 , 1 ) (0,1)(0,1)(0,1) est la limite d'une suite edy polynomes conservant les bornes, les variations et la classe de la fonction, convergeant uniformément dans tout l'intervalle.
4. L'ordre de l'approximation par les polynomes P n ( x ; f ) P n ( x ; f ) P_(n)(x;f)\mathrm{P}_{n}(x ; f)Pn(x;f) s'obtient facilement. Désignons par ω ( δ ) ω ( δ ) omega(delta)\omega(\delta)ω(δ) le module d'oscillation de la fonction f ( x ) ( 5 ) f ( x ) ( 5 ) f(x)(5)f(x)(5)f(x)(5). Nous avons
(3) | f ( x ) P n ( x ; f ) | i = 0 n | f ( x ) f ( i n ) | ( n i ) x t ( 1 x ) n t f ( x ) P n ( x ; f ) i = 0 n f ( x ) f i n ( n i ) x t ( 1 x ) n t |f(x)-P_(n)(x;f)| <= sum_(i=0)^(n)|f(x)-f((i)/(n))|((n)/(i))x^(t)(1-x)^(n-t) <=\left|f(x)-\mathrm{P}_{n}(x ; f)\right| \leq \sum_{i=0}^{n}\left|f(x)-f\left(\frac{i}{n}\right)\right|\binom{n}{i} x^{t}(1-x)^{n-t} \leq|f(x)Pn(x;f)|i=0n|f(x)f(in)|(ni)xt(1x)nt
i = 0 n ω ( | x i n | ) ( n i ) x i ( 1 x ) n i < < ( 1 δ i = 0 n | x i n | ( n i ) x i ( 1 x ) n i + 1 ) ω ( δ ) i = 0 n ω x i n ( n i ) x i ( 1 x ) n i < < 1 δ i = 0 n x i n ( n i ) x i ( 1 x ) n i + 1 ω ( δ ) {:[ <= sum_(i=0)^(n)omega(|x-(i)/(n)|)((n)/(i))x^(i)(1-x)^(n-i) < ],[ < ((1)/(delta)sum_(i=0)^(n)|x-(i)/(n)|((n)/(i))x^(i)(1-x)^(n-i)+1)omega(delta)]:}\begin{aligned} \leq \sum_{i=0}^{n} \omega & \left(\left|x-\frac{i}{n}\right|\right)\binom{n}{i} x^{i}(1-x)^{n-i}< \\ & <\left(\frac{1}{\delta} \sum_{i=0}^{n}\left|x-\frac{i}{n}\right|\binom{n}{i} x^{i}(1-x)^{n-i}+1\right) \omega(\delta) \end{aligned}i=0nω(|xin|)(ni)xi(1x)ni<<(1δi=0n|xin|(ni)xi(1x)ni+1)ω(δ)
Or dans l'intervalle ( j n , j + 1 n ) j n , j + 1 n ((j)/(n),(j+1)/(n))\left(\frac{j}{n}, \frac{j+1}{n}\right)(jn,j+1n) nous avons
i = 0 n | x i n | ( n i ) x i ( 1 x ) n i = 2 ( n 1 j ) x j + 1 ( 1 x ) n 1 2 ( n 1 j ) ( j n , j + 1 n ) x j + 1 ( 1 x ) n 1 = 2 ( n 1 i ) ( 1 + 1 ) j + 1 ( n j ) n j ( n + 1 ) n + 1 i = 0 n x i n n i x i ( 1 x ) n i = 2 ( n 1 j ) x j + 1 ( 1 x ) n 1 2 ( n 1 j ) j n , j + 1 n x j + 1 ( 1 x ) n 1 = 2 ( n 1 i ) ( 1 + 1 ) j + 1 ( n j ) n j ( n + 1 ) n + 1 {:[sum_(i=0)^(n)|x-(i)/(n)|((n)/(i))x^(i)(1-x)^(n-i)=2((n-1)/(j))x^(j+1)(1-x)^(n-1) <= ],[ <= 2((n-1)/(j))_(((j)/(n),(j+1)/(n)))x^(j+1)(1-x)^(n-1)=2((n-1)/(i))((1+1)^(j+1)(n-j)^(n-j))/((n+1)^(n+1))]:}\begin{aligned} & \sum_{i=0}^{n}\left|x-\frac{i}{n}\right|\left(\frac{n}{i}\right) x^{i}(1-x)^{n-i}=2\binom{n-1}{j} x^{j+1}(1-x)^{n-1} \leq \\ & \leq 2\binom{n-1}{j}_{\left(\frac{j}{n}, \frac{j+1}{n}\right)} x^{j+1}(1-x)^{n-1}=2\binom{n-1}{i} \frac{(1+1)^{j+1}(n-j)^{n-j}}{(n+1)^{n+1}} \end{aligned}i=0n|xin|(ni)xi(1x)ni=2(n1j)xj+1(1x)n12(n1j)(jn,j+1n)xj+1(1x)n1=2(n1i)(1+1)j+1(nj)nj(n+1)n+1
et on en déduit
max . ( 0 , 1 ) i = 0 n | x i n | ( n i ) x i ( 1 x ) n i = M n = { 2 ( n 1 n 2 ) ( n 2 + 1 ) n 2 + 1 ( n 2 ) n 2 n + 1 ) n + 1 , x pair 1 2 n ( n 1 n 1 2 ) , n impair. max . ( 0 , 1 ) i = 0 n x i n ( n i ) x i ( 1 x ) n i = M n = 2 ( n 1 n 2 ) n 2 + 1 n 2 + 1 n 2 n 2 n + 1 ) n + 1 , x  pair  1 2 n ( n 1 n 1 2 ) , n  impair.  max._((0,1))sum_(i=0)^(n)|x-(i)/(n)|((n)/(i))xi(1-x)^(n-i)=M_(n)={[2((n-1)/((n)/(2)))(((n)/(2)+1)^((n)/(2)+1)((n)/(2))^((n)/(2)))/(n+1)^(n+1))","x" pair "],[(1)/(2^(n))((n-1)/((n-1)/(2)))","n" impair. "]:}\operatorname{max.}_{(0,1)} \sum_{i=0}^{n}\left|x-\frac{i}{n}\right|\binom{n}{i} x i(1-x)^{n-i}=\mathrm{M}_{n}=\left\{\begin{array}{r}2\binom{n-1}{\frac{n}{2}} \frac{\left(\frac{n}{2}+1\right)^{\frac{n}{2}+1}\left(\frac{n}{2}\right)^{\frac{n}{2}}}{n+1)^{n+1}}, x \text { pair } \\ \frac{1}{2^{n}}\binom{n-1}{\frac{n-1}{2}}, n \text { impair. }\end{array}\right.max.(0,1)i=0n|xin|(ni)xi(1x)ni=Mn={2(n1n2)(n2+1)n2+1(n2)n2n+1)n+1,x pair 12n(n1n12),n impair. 
On voit immédiatement que
n 1 2 M n 1 2 π n 1 2 M n 1 2 π n^((1)/(2))M_(n)rarr(1)/(sqrt(2pi))n^{\frac{1}{2}} M_{n} \rightarrow \frac{1}{\sqrt{2 \pi}}n12Mn12π
et
( 2 n 1 ) 1 2 M 2 n + 1 > ( 2 n + 1 ) 1 2 M 2 n + 1 ( 2 n 1 ) 1 2 M 2 n + 1 > ( 2 n + 1 ) 1 2 M 2 n + 1 (2n-1)^((1)/(2))M_(2n+1) > (2n+1)^((1)/(2))M_(2n+1)(2 n-1)^{\frac{1}{2}} M_{2 n+1}>(2 n+1)^{\frac{1}{2}} M_{2 n+1}(2n1)12M2n+1>(2n+1)12M2n+1
»Leçons sur l'approximation des fonctions d'une variable réelle" p. 7.
độoù
M 1 = 1 2 , M 2 n + 1 3 4 2 n + 1 n = 1 , 2 , M 1 = 1 2 , M 2 n + 1 3 4 2 n + 1 n = 1 , 2 , M_(1)=(1)/(2),M_(2n+1) <= (sqrt3)/(4sqrt(2n+1))quad n=1,2,dots\mathrm{M}_{1}=\frac{1}{2}, \mathrm{M}_{2 n+1} \leq \frac{\sqrt{3}}{4 \sqrt{2 n+1}} \quad n=1,2, \ldotsM1=12,M2n+1342n+1n=1,2,
Si n n nnn est pair ( n 2 n 2 n >= 2n \geq 2n2 ) on établit facilement que
M n < M n + 1 n + 2 n 3 4 n + 1 n + 2 n 1 2 n . M n < M n + 1 n + 2 n 3 4 n + 1 n + 2 n 1 2 n . M_(n) < M_(n+1)sqrt((n+2)/(n)) <= (sqrt3)/(4sqrt(n+1))sqrt((n+2)/(n)) <= (1)/(2sqrtn).\mathrm{M}_{n}<\mathrm{M}_{n+1} \sqrt{\frac{n+2}{n}} \leq \frac{\sqrt{3}}{4 \sqrt{n+1}} \sqrt{\frac{n+2}{n}} \leq \frac{1}{2 \sqrt{n}} .Mn<Mn+1n+2n34n+1n+2n12n.
Donc quel que soit n n nnn
M n 1 2 n M n 1 2 n M_(n) <= (1)/(2sqrtn)\mathrm{M}_{n} \leq \frac{1}{2 \sqrt{n}}Mn12n
Faisant δ = 1 n δ = 1 n delta=(1)/(sqrtn)\delta=\frac{1}{\sqrt{n}}δ=1n dans la formule (3) nous avons
| f ( x ) P n ( x ; f ) | < 3 2 ω ( 1 n ) dans ( 0 , 1 ) ( 6 ) f ( x ) P n ( x ; f ) < 3 2 ω 1 n  dans  ( 0 , 1 ) ( 6 ) |f(x)-P_(n)(x;f)| < (3)/(2)omega((1)/(sqrtn))quad" dans "(0,1)(6)\left|f(x)-\mathrm{P}_{n}(x ; f)\right|<\frac{3}{2} \omega\left(\frac{1}{\sqrt{n}}\right) \quad \text { dans }(0,1)(6)|f(x)Pn(x;f)|<32ω(1n) dans (0,1)(6)
L'approximation et en général effectivement de l'ordre de ω ( 1 n ) ω 1 n omega((1)/(sqrtn))\omega\left(\frac{1}{\sqrt{n}}\right)ω(1n). Soit en effet la fonction f ( x ) = | 1 2 x | f ( x ) = 1 2 x f(x)=|(1)/(2)-x|f(x)=\left|\frac{1}{2}-x\right|f(x)=|12x|. On a dans ce cas ω ( δ ) = δ ω ( δ ) = δ omega(delta)=delta\omega(\delta)=\deltaω(δ)=δ pour δ 1 2 δ 1 2 delta <= (1)/(2)\delta \leq \frac{1}{2}δ12 et
P 2 n + 1 ( 1 2 ; f ) f ( 1 2 ) = 1 2 2 n + 1 ( 2 n n ) n > 1 P 2 n + 1 1 2 ; f f 1 2 = 1 2 2 n + 1 ( 2 n n ) n > 1 P_(2n+1)((1)/(2);f)-f((1)/(2))=(1)/(2^(2n+1))((2n)/(n))quad n > 1\mathrm{P}_{2 n+1}\left(\frac{1}{2} ; f\right)-f\left(\frac{1}{2}\right)=\frac{1}{2^{2 n+1}}\binom{2 n}{n} \quad n>1P2n+1(12;f)f(12)=122n+1(2nn)n>1
d'où
max . | f ( x ) P 2 n + 1 ( x ; f ) | 1 2 π 1 2 n + 1 = 1 2 π ω ( 1 2 n + 1 ) ( 7 ) max . f ( x ) P 2 n + 1 ( x ; f ) 1 2 π 1 2 n + 1 = 1 2 π ω 1 2 n + 1 ( 7 ) max.|f(x)-P_(2n+1)(x;f)| >= (1)/(sqrt(2pi))*(1)/(sqrt(2n+1))=(1)/(sqrt(2pi))^(omega)((1)/(sqrt(2n+1)))^((7))\operatorname{max.}\left|f(x)-\mathrm{P}_{2 n+1}(x ; f)\right| \geq \frac{1}{\sqrt{2 \pi}} \cdot \frac{1}{\sqrt{2 n+1}}=\frac{1}{\sqrt{2 \pi}}{ }^{\omega}\left(\frac{1}{\sqrt{2 n+1}}\right){ }^{(7)}max.|f(x)P2n+1(x;f)|12π12n+1=12πω(12n+1)(7).
( 6 6 ^(6){ }^{6}6 ) La démonstration donnée dans le livre cité de MM. Pólya et Szegö me donne que
| f ( x ) P n ( x ; f ) | < ω ( 1 4 n ) + Δ 0 [ f ] 2 n f ( x ) P n ( x ; f ) < ω 1 4 n + Δ 0 [ f ] 2 n |f(x)-P_(n)(x;f)| < omega((1)/((4)/(sqrtn)))+(Delta_(0)[f])/(2sqrtn)\left|f(x)-\mathrm{P}_{n}(x ; f)\right|<\omega\left(\frac{1}{\frac{4}{\sqrt{n}}}\right)+\frac{\Delta_{0}[f]}{2 \sqrt{n}}|f(x)Pn(x;f)|<ω(14n)+Δ0[f]2n
Si la fonction vérifie une condition de Lipschitz ordinaire on a
| t ( x ) P n ( x ; f ) | < 3 2 Δ 1 [ f ] n . t ( x ) P n ( x ; f ) < 3 2 Δ 1 [ f ] n . |t(x)-P_(n)(x;f)| < (3)/(2)(Delta_(1)[f])/(sqrtn).\left|t(x)-P_{n}(x ; f)\right|<\frac{3}{2} \frac{\Delta_{1}[f]}{\sqrt{n}} .|t(x)Pn(x;f)|<32Δ1[f]n.
M. Wigert ne donne que l'approximation
3 2 Δ 0 [ f ] Δ 1 [ f ] 2 3 1 n 3 . 3 2 Δ 0 [ f ] Δ 1 [ f ] 2 3 1 n 3 . (3)/(2)root(3)(Delta_(0)[f]*Delta_(1)[f]^(2))*(1)/(root(3)(n)).\frac{3}{2} \sqrt[3]{\Delta_{0}[f] \cdot \Delta_{1}[f]^{2}} \cdot \frac{1}{\sqrt[3]{n}} .32Δ0[f]Δ1[f]231n3.
(7) Nons avons supposé que la fonction est définie dans l'intervalle ( 0 , 1 ) ( 0 , 1 ) (0,1)(0,1)(0,1) uniquement pour simplifier l'éxposé. On aurrait pu considérer des fonc-

  1. (5) Pour les propriétés de ω ( δ ) ω ( δ ) omega(delta)\omega(\delta)ω(δ) voir par ex. C H C H C_(H)\mathrm{C}_{\mathrm{H}}CH. de la Vallee Poussin
1935

Related Posts