Namespaces
Variants
Actions

Difference between revisions of "Absolutely summing operator"

From Encyclopedia of Mathematics
Jump to: navigation, search
(Importing text file)
 
m (tex encoded by computer)
 
Line 1: Line 1:
A [[Linear operator|linear operator]] <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/a/a110/a110070/a1100701.png" /> acting from a [[Banach space|Banach space]] <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/a/a110/a110070/a1100702.png" /> into a Banach space <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/a/a110/a110070/a1100703.png" /> is called absolutely <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/a/a110/a110070/a1100705.png" />-summing (<img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/a/a110/a110070/a1100706.png" />) if there is a constant <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/a/a110/a110070/a1100707.png" /> such that
+
<!--
 +
a1100701.png
 +
$#A+1 = 36 n = 0
 +
$#C+1 = 36 : ~/encyclopedia/old_files/data/A110/A.1100070 Absolutely summing operator
 +
Automatically converted into TeX, above some diagnostics.
 +
Please remove this comment and the {{TEX|auto}} line below,
 +
if TeX found to be correct.
 +
-->
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/a/a110/a110070/a1100708.png" /></td> </tr></table>
+
{{TEX|auto}}
 +
{{TEX|done}}
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/a/a110/a110070/a1100709.png" /></td> </tr></table>
+
A [[Linear operator|linear operator]]  $  T $
 +
acting from a [[Banach space|Banach space]]  $  X $
 +
into a Banach space  $  Y $
 +
is called absolutely  $  p $-
 +
summing ( $  1 \leq  p < \infty $)
 +
if there is a constant  $  c \geq 0 $
 +
such that
  
whenever <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/a/a110/a110070/a11007010.png" /> and <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/a/a110/a110070/a11007011.png" />. Here, <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/a/a110/a110070/a11007012.png" /> denotes the value of the [[Linear functional|linear functional]] <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/a/a110/a110070/a11007013.png" /> (the Banach dual of <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/a/a110/a110070/a11007014.png" />, cf. [[Adjoint space|Adjoint space]]) at the element <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/a/a110/a110070/a11007015.png" />. The set of such operators, denoted by <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/a/a110/a110070/a11007016.png" />, becomes a Banach space under the norm <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/a/a110/a110070/a11007018.png" />, and <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/a/a110/a110070/a11007019.png" /> is a Banach operator ideal. If <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/a/a110/a110070/a11007020.png" />, then <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/a/a110/a110070/a11007021.png" />.
+
$$
 +
\left ( \sum _ {k =1 } ^ { n }  \left \| {Tx _ {k} } \right \| ^ {p} \right ) ^ {1/p } \leq
 +
$$
  
The prototype of an absolutely <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/a/a110/a110070/a11007022.png" />-summing operator is the canonical mapping <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/a/a110/a110070/a11007023.png" />, where <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/a/a110/a110070/a11007024.png" /> is a [[Borel measure|Borel measure]] on a compact [[Hausdorff space|Hausdorff space]] <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/a/a110/a110070/a11007025.png" />. In this case, <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/a/a110/a110070/a11007026.png" />.
+
$$
 +
\leq 
 +
c  \sup  \left \{ {\left ( \sum _ {k = 1 } ^ { n }  \left | {\left \langle  {x _ {k} ,a } \right \rangle } \right |  ^ {p} \right ) ^ {1/p } } : {a \in X  ^  \prime  , \left \| a \right \| \leq 1 } \right \}
 +
$$
  
The famous Grothendieck theorem says that all operators from <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/a/a110/a110070/a11007027.png" /> into any [[Hilbert space|Hilbert space]] are absolutely <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/a/a110/a110070/a11007028.png" />-summing.
+
whenever  $  x _ {1} \dots x _ {n} \in X $
 +
and  $  n = 1,2, \dots $.  
 +
Here,  $  \langle  {x _ {k} ,a } \rangle $
 +
denotes the value of the [[Linear functional|linear functional]]  $  a \in X  ^  \prime  $(
 +
the Banach dual of  $  X $,
 +
cf. [[Adjoint space|Adjoint space]]) at the element  $  x _ {k} \in X $.
 +
The set of such operators, denoted by  $  \Pi _ {p} ( X,Y ) $,
 +
becomes a Banach space under the norm  $  \pi _ {p} ( T ) = \inf  c $,
 +
and  $  \Pi _ {p} = \cup _ {X,Y }  \Pi _ {p} ( X,Y ) $
 +
is a Banach operator ideal. If  $  1 \leq  p < q < \infty $,
 +
then  $  \Pi _ {p} \subset  \Pi _ {q} $.
  
Absolutely <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/a/a110/a110070/a11007029.png" />-summing operators are weakly compact but may fail to be compact (cf. also [[Compact operator|Compact operator]]). For a Hilbert space <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/a/a110/a110070/a11007030.png" /> it turns out that <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/a/a110/a110070/a11007031.png" /> is just the set of Hilbert–Schmidt operators (cf. [[Hilbert–Schmidt operator|Hilbert–Schmidt operator]]). Nuclear operators (cf. [[Nuclear operator|Nuclear operator]]) are absolutely <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/a/a110/a110070/a11007032.png" />-summing. Conversely, the product of <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/a/a110/a110070/a11007033.png" /> absolutely <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/a/a110/a110070/a11007034.png" />-summing operators is nuclear, hence compact, if <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/a/a110/a110070/a11007035.png" />. This implies that the identity mapping of a Banach space <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/a/a110/a110070/a11007036.png" /> is absolutely <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/a/a110/a110070/a11007037.png" />-summing if and only if <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/a/a110/a110070/a11007038.png" /> (the Dvoretzky–Rogers theorem).
+
The prototype of an absolutely  $  p $-
 +
summing operator is the canonical mapping  $  { { \mathop{\rm Id} } } : {C ( K ) } \rightarrow {L _ {p} ( K, \mu ) } $,
 +
where  $  \mu $
 +
is a [[Borel measure|Borel measure]] on a compact [[Hausdorff space|Hausdorff space]]  $  K $.
 +
In this case,  $  \pi _ {p} ( { { \mathop{\rm Id} } } : {C ( K ) } \rightarrow {L _ {p} ( K, \mu ) } ) = \mu ( K ) ^ {1/p } $.
 +
 
 +
The famous Grothendieck theorem says that all operators from  $  L _ {1} ( K, \mu ) $
 +
into any [[Hilbert space|Hilbert space]] are absolutely  $  1 $-
 +
summing.
 +
 
 +
Absolutely  $  p $-
 +
summing operators are weakly compact but may fail to be compact (cf. also [[Compact operator|Compact operator]]). For a Hilbert space $  H $
 +
it turns out that $  \Pi _ {p} ( H,H ) $
 +
is just the set of Hilbert–Schmidt operators (cf. [[Hilbert–Schmidt operator|Hilbert–Schmidt operator]]). Nuclear operators (cf. [[Nuclear operator|Nuclear operator]]) are absolutely $  p $-
 +
summing. Conversely, the product of $  2n $
 +
absolutely $  p $-
 +
summing operators is nuclear, hence compact, if $  2n \geq  p $.  
 +
This implies that the identity mapping of a Banach space $  X $
 +
is absolutely $  p $-
 +
summing if and only if $  { \mathop{\rm dim} } ( X ) < \infty $(
 +
the Dvoretzky–Rogers theorem).
  
 
====References====
 
====References====
 
<table><TR><TD valign="top">[a1]</TD> <TD valign="top">  J. Diestel,  H. Jarchow,  A. Tonge,  "Absolutely summing operators" , Cambridge Univ. Press  (1995)</TD></TR><TR><TD valign="top">[a2]</TD> <TD valign="top">  G.J.O. Jameson,  "Summing and nuclear norms in Banach space theory" , Cambridge Univ. Press  (1987)</TD></TR><TR><TD valign="top">[a3]</TD> <TD valign="top">  A. Pietsch,  "Operator ideals" , North-Holland  (1980)</TD></TR></table>
 
<table><TR><TD valign="top">[a1]</TD> <TD valign="top">  J. Diestel,  H. Jarchow,  A. Tonge,  "Absolutely summing operators" , Cambridge Univ. Press  (1995)</TD></TR><TR><TD valign="top">[a2]</TD> <TD valign="top">  G.J.O. Jameson,  "Summing and nuclear norms in Banach space theory" , Cambridge Univ. Press  (1987)</TD></TR><TR><TD valign="top">[a3]</TD> <TD valign="top">  A. Pietsch,  "Operator ideals" , North-Holland  (1980)</TD></TR></table>

Latest revision as of 16:08, 1 April 2020


A linear operator $ T $ acting from a Banach space $ X $ into a Banach space $ Y $ is called absolutely $ p $- summing ( $ 1 \leq p < \infty $) if there is a constant $ c \geq 0 $ such that

$$ \left ( \sum _ {k =1 } ^ { n } \left \| {Tx _ {k} } \right \| ^ {p} \right ) ^ {1/p } \leq $$

$$ \leq c \sup \left \{ {\left ( \sum _ {k = 1 } ^ { n } \left | {\left \langle {x _ {k} ,a } \right \rangle } \right | ^ {p} \right ) ^ {1/p } } : {a \in X ^ \prime , \left \| a \right \| \leq 1 } \right \} $$

whenever $ x _ {1} \dots x _ {n} \in X $ and $ n = 1,2, \dots $. Here, $ \langle {x _ {k} ,a } \rangle $ denotes the value of the linear functional $ a \in X ^ \prime $( the Banach dual of $ X $, cf. Adjoint space) at the element $ x _ {k} \in X $. The set of such operators, denoted by $ \Pi _ {p} ( X,Y ) $, becomes a Banach space under the norm $ \pi _ {p} ( T ) = \inf c $, and $ \Pi _ {p} = \cup _ {X,Y } \Pi _ {p} ( X,Y ) $ is a Banach operator ideal. If $ 1 \leq p < q < \infty $, then $ \Pi _ {p} \subset \Pi _ {q} $.

The prototype of an absolutely $ p $- summing operator is the canonical mapping $ { { \mathop{\rm Id} } } : {C ( K ) } \rightarrow {L _ {p} ( K, \mu ) } $, where $ \mu $ is a Borel measure on a compact Hausdorff space $ K $. In this case, $ \pi _ {p} ( { { \mathop{\rm Id} } } : {C ( K ) } \rightarrow {L _ {p} ( K, \mu ) } ) = \mu ( K ) ^ {1/p } $.

The famous Grothendieck theorem says that all operators from $ L _ {1} ( K, \mu ) $ into any Hilbert space are absolutely $ 1 $- summing.

Absolutely $ p $- summing operators are weakly compact but may fail to be compact (cf. also Compact operator). For a Hilbert space $ H $ it turns out that $ \Pi _ {p} ( H,H ) $ is just the set of Hilbert–Schmidt operators (cf. Hilbert–Schmidt operator). Nuclear operators (cf. Nuclear operator) are absolutely $ p $- summing. Conversely, the product of $ 2n $ absolutely $ p $- summing operators is nuclear, hence compact, if $ 2n \geq p $. This implies that the identity mapping of a Banach space $ X $ is absolutely $ p $- summing if and only if $ { \mathop{\rm dim} } ( X ) < \infty $( the Dvoretzky–Rogers theorem).

References

[a1] J. Diestel, H. Jarchow, A. Tonge, "Absolutely summing operators" , Cambridge Univ. Press (1995)
[a2] G.J.O. Jameson, "Summing and nuclear norms in Banach space theory" , Cambridge Univ. Press (1987)
[a3] A. Pietsch, "Operator ideals" , North-Holland (1980)
How to Cite This Entry:
Absolutely summing operator. Encyclopedia of Mathematics. URL: http://encyclopediaofmath.org/index.php?title=Absolutely_summing_operator&oldid=18591
This article was adapted from an original article by A. Pietsch (originator), which appeared in Encyclopedia of Mathematics - ISBN 1402006098. See original article