Beruflich Dokumente
Kultur Dokumente
In der Mathematik versteht man unter einer Matrix (Plural Matrizen) eine
rechteckige Anordnung (Tabelle) von Elementen (meist mathematischer Objekte, etwa
Zahlen). Mit diesen Objekten lässt sich dann in bestimmter Weise rechnen, indem man
Matrizen addiert oder miteinander multipliziert.
Matrizen sind ein Schlüsselkonzept der linearen Algebra und tauchen in fast allen
Gebieten der Mathematik auf. Sie stellen Zusammenhänge, in denen
Linearkombinationen eine Rolle spielen, übersichtlich dar und erleichtern damit
Rechen- und Gedankenvorgänge. Sie werden insbesondere dazu benutzt, lineare
Abbildungen darzustellen und lineare Gleichungssysteme zu beschreiben und zu lösen.
Die Bezeichnung Matrix wurde 1850 von James Joseph Sylvester eingeführt.
Als Notation hat sich die Anordnung der Elemente in Zeilen und Spalten zwischen
zwei großen öffnenden und schließenden Klammern durchgesetzt. In der Regel
verwendet man runde Klammern, es werden aber auch eckige verwendet. Zum Beispiel
bezeichnen
( a 11 a 12 a 13 a 21 a 22 a 23 ) {\displaystyle
{\begin{pmatrix}a_{11}&a_{12}&a_{13}\\a_{21}&a_{22}&a_{23}\end{pmatrix}}}
{\displaystyle
{\begin{pmatrix}a_{11}&a_{12}&a_{13}\\a_{21}&a_{22}&a_{23}\end{pmatrix}}} und [ a
11 a 12 a 13 a 21 a 22 a 23 ] {\displaystyle
{\begin{bmatrix}a_{11}&a_{12}&a_{13}\\a_{21}&a_{22}&a_{23}\end{bmatrix}}}
{\displaystyle
{\begin{bmatrix}a_{11}&a_{12}&a_{13}\\a_{21}&a_{22}&a_{23}\end{bmatrix}}}
Matrizen mit zwei Zeilen und drei Spalten. Matrizen werden üblicherweise mit
Großbuchstaben (manchmal fett gedruckt oder, handschriftlich, einfach oder doppelt
unterstrichen), vorzugsweise A {\displaystyle A} A, bezeichnet. Eine Matrix mit m
{\displaystyle m} m Zeilen und n {\displaystyle n} n Spalten:
A = A = A _ = ( a 11 a 12 ⋯ a 1 n a 21 a 22 ⋯ a 2 n ⋮ ⋮ ⋮ a m 1 a m 2 ⋯ a m n )
= ( a i j ) {\displaystyle A={\boldsymbol {A}}={\underline
{A}}={\begin{pmatrix}a_{11}&a_{12}&\cdots &a_{1n}\\a_{21}&a_{22}&\cdots
&a_{2n}\\\vdots &\vdots &&\vdots \\a_{m1}&a_{m2}&\cdots
&a_{mn}\\\end{pmatrix}}=(a_{ij})} {\displaystyle A={\boldsymbol {A}}={\underline
{A}}={\begin{pmatrix}a_{11}&a_{12}&\cdots &a_{1n}\\a_{21}&a_{22}&\cdots
&a_{2n}\\\vdots &\vdots &&\vdots \\a_{m1}&a_{m2}&\cdots
&a_{mn}\\\end{pmatrix}}=(a_{ij})}.
Die Elemente der Matrix nennt man auch Einträge oder Komponenten der Matrix. Sie
entstammen einer Menge K , {\displaystyle K,} K, in der Regel einem Körper oder
einem Ring. Man spricht von einer Matrix über K {\displaystyle K} K. Wählt man für
K {\displaystyle K} K die Menge der reellen Zahlen, so spricht man von einer
reellen Matrix, bei komplexen Zahlen von einer komplexen Matrix.
Ein bestimmtes Element beschreibt man durch zwei Indizes, meist ist das Element in
der ersten Zeile und der ersten Spalte durch a 11 {\displaystyle a_{11}} a_{11}
beschrieben. Allgemein bezeichnet a i j {\displaystyle a_{ij}} a_{ij} das Element
in der i {\displaystyle i} i-ten Zeile und der j {\displaystyle j} j-ten Spalte.
Bei der Indizierung wird dabei stets als erstes der Zeilenindex und als zweites der
Spaltenindex des Elements genannt. Merkregel: Zeile zuerst, Spalte später. Wenn
Verwechslungsgefahr besteht, werden die beiden Indizes mit einem Komma abgetrennt.
So wird zum Beispiel das Matrixelement in der ersten Zeile und der elften Spalte
mit a 1 , 11 {\displaystyle a_{1,11}} a_{1,11} bezeichnet.
Einzelne Zeilen und Spalten werden oft als Spalten- oder Zeilenvektoren bezeichnet.
Ein Beispiel:
A = ( a 11 a 12 a 21 a 22 ) , {\displaystyle
A={\begin{pmatrix}a_{11}&a_{12}\\a_{21}&a_{22}\end{pmatrix}},} A = \begin{pmatrix}
a_{11} & a_{12} \\ a_{21} & a_{22} \end{pmatrix}, hier sind ( a 11 a 21 )
{\displaystyle {\begin{pmatrix}a_{11}\\a_{21}\end{pmatrix}}} \begin{pmatrix} a_{11}
\\ a_{21} \end{pmatrix} und ( a 12 a 22 ) {\displaystyle
{\begin{pmatrix}a_{12}\\a_{22}\end{pmatrix}}} \begin{pmatrix} a_{12} \\ a_{22}
\end{pmatrix} die Spalten oder Spaltenvektoren sowie ( a 11 a 12 ) {\displaystyle
{\begin{pmatrix}a_{11}&a_{12}\end{pmatrix}}} \begin{pmatrix} a_{11} & a_{12}
\end{pmatrix} und ( a 21 a 22 ) {\displaystyle
{\begin{pmatrix}a_{21}&a_{22}\end{pmatrix}}} \begin{pmatrix} a_{21} & a_{22}
\end{pmatrix} die Zeilen oder Zeilenvektoren.
Bei einzeln stehenden Zeilen- und Spaltenvektoren einer Matrix wird gelegentlich
der unveränderliche Index weggelassen. Manchmal werden Spaltenvektoren zur
kompakteren Darstellung als transponierte Zeilenvektoren geschrieben, also:
( a 11 a 21 ) {\displaystyle {\begin{pmatrix}a_{11}\\a_{21}\end{pmatrix}}}
\begin{pmatrix} a_{11} \\ a_{21} \end{pmatrix} oder ( a 1 a 2 ) {\displaystyle
{\begin{pmatrix}a_{1}\\a_{2}\end{pmatrix}}} {\displaystyle
{\begin{pmatrix}a_{1}\\a_{2}\end{pmatrix}}} als ( a 11 a 21 ) T {\displaystyle
{\begin{pmatrix}a_{11}&a_{21}\end{pmatrix}}^{T}} {\displaystyle
{\begin{pmatrix}a_{11}&a_{21}\end{pmatrix}}^{T}} oder ( a 1 a 2 ) T {\displaystyle
{\begin{pmatrix}a_{1}&a_{2}\end{pmatrix}}^{T}} {\displaystyle
{\begin{pmatrix}a_{1}&a_{2}\end{pmatrix}}^{T}}
Typ
Der Typ einer Matrix ergibt sich aus der Anzahl ihrer Zeilen und Spalten. Eine
Matrix mit m {\displaystyle m} m Zeilen und n {\displaystyle n} n Spalten nennt man
eine m × n {\displaystyle m\times n} m\times n-Matrix (sprich: m-mal-n- oder m-
Kreuz-n-Matrix). Stimmen Zeilen- und Spaltenanzahl überein, so spricht man von
einer quadratischen Matrix.
Eine Matrix, die aus nur einer Spalte oder nur einer Zeile besteht, wird
üblicherweise als Vektor aufgefasst. Einen Vektor mit n {\displaystyle n} n
Elementen kann man je nach Kontext als einspaltige n × 1 {\displaystyle n\times 1}
n \times 1-Matrix oder einzeilige 1 × n {\displaystyle 1\times n} 1 \times n-Matrix
darstellen. Neben den Begriffen Spaltenvektor und Zeilenvektor sind hierfür auch
die Begriffe Spaltenmatrix und Zeilenmatrix geläufig. Eine 1 × 1 {\displaystyle
1\times 1} 1\times 1-Matrix ist sowohl Spalten- als auch Zeilenmatrix und wird als
Skalar angesehen.
Formale Darstellung
Eine Matrix ist eine doppelt indizierte Familie. Formal ist dies eine Funktion
A : { 1 , … , m } × { 1 , … , n } → K , ( i , j ) ↦ a i j , {\displaystyle
A\colon \{1,\dotsc ,m\}\times \{1,\dotsc ,n\}\to K,\quad (i,j)\mapsto a_{ij},}
{\displaystyle A\colon \{1,\dotsc ,m\}\times \{1,\dotsc ,n\}\to K,\quad
(i,j)\mapsto a_{ij},}
Zwei Matrizen können addiert werden, wenn sie vom selben Typ sind, das heißt, wenn
sie dieselbe Anzahl von Zeilen und dieselbe Anzahl von Spalten besitzen. Die Summe
zweier m × n {\displaystyle m\times n} m\times n-Matrizen ist komponentenweise
definiert:
A + B := ( a i j + b i j ) i = 1 , … , m ; j = 1 , … , n {\displaystyle
A+B:=(a_{ij}+b_{ij})_{i=1,\dotsc ,m;\ j=1,\dotsc ,n}} {\displaystyle A+B:=(a_{ij}
+b_{ij})_{i=1,\dotsc ,m;\ j=1,\dotsc ,n}}
Rechenbeispiel:
( 1 − 3 2 1 2 7 ) + ( 0 3 5 2 1 − 1 ) = ( 1 + 0 − 3 + 3 2 + 5 1 + 2 2 + 1 7 + (
− 1 ) ) = ( 1 0 7 3 3 6 ) {\displaystyle {\begin{pmatrix}1&-
3&2\\1&2&7\end{pmatrix}}+{\begin{pmatrix}0&3&5\\2&1&-
1\end{pmatrix}}={\begin{pmatrix}1+0&-3+3&2+5\\1+2&2+1&7+(-
1)\end{pmatrix}}={\begin{pmatrix}1&0&7\\3&3&6\end{pmatrix}}} {\displaystyle
{\begin{pmatrix}1&-3&2\\1&2&7\end{pmatrix}}+{\begin{pmatrix}0&3&5\\2&1&-
1\end{pmatrix}}={\begin{pmatrix}1+0&-3+3&2+5\\1+2&2+1&7+(-
1)\end{pmatrix}}={\begin{pmatrix}1&0&7\\3&3&6\end{pmatrix}}}
In der linearen Algebra sind die Einträge der Matrizen üblicherweise Elemente eines
Körpers, wie z. B. der reellen oder komplexen Zahlen. In diesem Fall ist die
Matrizenaddition assoziativ, kommutativ und besitzt mit der Nullmatrix ein
neutrales Element. Im Allgemeinen besitzt die Matrizenaddition diese Eigenschaften
jedoch nur, wenn die Einträge Elemente einer algebraischen Struktur sind, die diese
Eigenschaften hat.
Skalarmultiplikation
→ Hauptartikel: Skalarmultiplikation
Eine Matrix wird mit einem Skalar multipliziert, indem jeder Eintrag der Matrix mit
dem Skalar multipliziert wird:
λ ⋅ A := ( λ ⋅ a i j ) i = 1 , … , m ; j = 1 , … , n {\displaystyle
\lambda \cdot A:=(\lambda \cdot a_{ij})_{i=1,\dotsc ,m;\ j=1,\dotsc ,n}}
{\displaystyle \lambda \cdot A:=(\lambda \cdot a_{ij})_{i=1,\dotsc ,m;\
j=1,\dotsc ,n}}
Rechenbeispiel:
5 ⋅ ( 1 − 3 2 1 2 7 ) = ( 5 ⋅ 1 5 ⋅ ( − 3 ) 5 ⋅ 2 5 ⋅ 1 5 ⋅ 2 5 ⋅ 7 ) = ( 5 −
15 10 5 10 35 ) {\displaystyle 5\cdot {\begin{pmatrix}1&-
3&2\\1&2&7\end{pmatrix}}={\begin{pmatrix}5\cdot 1&5\cdot (-3)&5\cdot 2\\5\cdot
1&5\cdot 2&5\cdot 7\end{pmatrix}}={\begin{pmatrix}5&-15&10\\5&10&35\end{pmatrix}}}
{\displaystyle 5\cdot {\begin{pmatrix}1&-
3&2\\1&2&7\end{pmatrix}}={\begin{pmatrix}5\cdot 1&5\cdot (-3)&5\cdot 2\\5\cdot
1&5\cdot 2&5\cdot 7\end{pmatrix}}={\begin{pmatrix}5&-15&10\\5&10&35\end{pmatrix}}}
Zwei Matrizen können multipliziert werden, wenn die Spaltenanzahl der linken mit
der Zeilenanzahl der rechten Matrix übereinstimmt. Das Produkt einer l × m
{\displaystyle l\times m} l \times m-Matrix A = ( a i j ) i = 1 , … , l , j = 1 , …
, m {\displaystyle A=(a_{ij})_{i=1,\dotsc ,l,\;j=1,\dotsc ,m}} {\displaystyle
A=(a_{ij})_{i=1,\dotsc ,l,\;j=1,\dotsc ,m}} und einer m × n {\displaystyle m\times
n} m\times n-Matrix B = ( b i j ) i = 1 , … , m , j = 1 , … , n {\displaystyle
B=(b_{ij})_{i=1,\dotsc ,m,\;j=1,\dotsc ,n}} {\displaystyle
B=(b_{ij})_{i=1,\dotsc ,m,\;j=1,\dotsc ,n}} ist eine l × n {\displaystyle l\times
n} l \times n-Matrix C = ( c i j ) i = 1 , … , l , j = 1 , … , n , {\displaystyle
C=(c_{ij})_{i=1,\dotsc ,l,\;j=1,\dotsc ,n},} {\displaystyle
C=(c_{ij})_{i=1,\dotsc ,l,\;j=1,\dotsc ,n},} deren Einträge berechnet werden, indem
die Produktsummenformel, ähnlich dem Skalarprodukt, auf Paare aus einem
Zeilenvektor der ersten und einem Spaltenvektor der zweiten Matrix angewandt wird:
c i j = ∑ k = 1 m a i k ⋅ b k j {\displaystyle c_{ij}=\sum
_{k=1}^{m}a_{ik}\cdot b_{kj}} {\displaystyle c_{ij}=\sum _{k=1}^{m}a_{ik}\cdot
b_{kj}}
A = ( a 11 … a 1 n ⋮ ⋮ a m 1 … a m n ) {\displaystyle
A={\begin{pmatrix}a_{11}&\dots &a_{1n}\\\vdots &&\vdots \\a_{m1}&\dots
&a_{mn}\end{pmatrix}}} A={\begin{pmatrix}a_{{11}}&\dots &a_{{1n}}\\\vdots
&&\vdots \\a_{{m1}}&\dots &a_{{mn}}\end{pmatrix}}
ist
A T = ( a 11 … a m 1 ⋮ ⋮ a 1 n … a m n ) {\displaystyle
A^{T}={\begin{pmatrix}a_{11}&\dots &a_{m1}\\\vdots &&\vdots \\a_{1n}&\dots
&a_{mn}\end{pmatrix}}} A^{T}={\begin{pmatrix}a_{{11}}&\dots &a_{{m1}}\\\vdots
&&\vdots \\a_{{1n}}&\dots &a_{{mn}}\end{pmatrix}}
die Transponierte. Man schreibt also die erste Zeile als erste Spalte, die zweite
Zeile als zweite Spalte usw. Die Matrix wird an ihrer Hauptdiagonalen a 11 , a 22 ,
… {\displaystyle a_{11},a_{22},\dotsc } {\displaystyle a_{11},a_{22},\dotsc }
gespiegelt. Es gelten die folgenden Rechenregeln:
( A + B ) T = A T + B T ( c ⋅ A ) T = c ⋅ A T ( A T ) T = A ( A ⋅ B ) T = B T ⋅
A T ( A − 1 ) T = ( A T ) − 1 {\displaystyle {\begin{aligned}(A+B)^{T}&=A^{T}
+B^{T}\\(c\cdot A)^{T}&=c\cdot A^{T}\\\left(A^{T}\right)^{T}&=A\\(A\cdot
B)^{T}&=B^{T}\cdot A^{T}\\\left(A^{-1}\right)^{T}&=\left(A^{T}\right)^{-
1}\end{aligned}}} {\displaystyle {\begin{aligned}(A+B)^{T}&=A^{T}+B^{T}\\(c\cdot
A)^{T}&=c\cdot A^{T}\\\left(A^{T}\right)^{T}&=A\\(A\cdot B)^{T}&=B^{T}\cdot
A^{T}\\\left(A^{-1}\right)^{T}&=\left(A^{T}\right)^{-1}\end{aligned}}}
Bei Matrizen über R {\displaystyle \mathbb {R} } \mathbb {R} ist die adjungierte
Matrix genau die transponierte Matrix.
Inverse Matrix
→ Hauptartikel: Inverse Matrix
der Beträge der beiden Vektoren und des Kosinus des von den beiden Vektoren
eingeschlossenen Winkels. Beispielsweise gilt
( 1 2 3 ) T ⋅ ( − 2 − 1 1 ) = ( 1 2 3 ) ⋅ ( − 2 − 1 1 ) = 1 ⋅ ( − 2 ) + 2 ⋅ ( −
1 ) + 3 ⋅ 1 = − 1 {\displaystyle {\begin{pmatrix}1\\2\\3\end{pmatrix}}^{T}\cdot
{\begin{pmatrix}-2\\-1\\1\end{pmatrix}}={\begin{pmatrix}1&2&3\end{pmatrix}}\cdot
{\begin{pmatrix}-2\\-1\\1\end{pmatrix}}=1\cdot (-2)+2\cdot (-1)+3\cdot 1=-1}
{\displaystyle {\begin{pmatrix}1\\2\\3\end{pmatrix}}^{T}\cdot {\begin{pmatrix}-2\\-
1\\1\end{pmatrix}}={\begin{pmatrix}1&2&3\end{pmatrix}}\cdot {\begin{pmatrix}-2\\-
1\\1\end{pmatrix}}=1\cdot (-2)+2\cdot (-1)+3\cdot 1=-1}
Das zweite Produkt v ⋅ w T {\displaystyle v\cdot w^{T}} v \cdot w^T ist eine n × n
{\displaystyle n\times n} n\times n-Matrix und heißt dyadisches Produkt oder
Tensorprodukt von v {\displaystyle v} v und w {\displaystyle w} w (geschrieben v ⊗
w {\displaystyle v\otimes w} v \otimes w). Seine Spalten sind skalare Vielfache von
v {\displaystyle v} v, seine Zeilen skalare Vielfache von w T {\displaystyle w^{T}}
w^T. Beispielsweise gilt
( 1 2 3 ) ⋅ ( − 2 − 1 1 ) T = ( 1 2 3 ) ⋅ ( − 2 − 1 1 ) = ( 1 ⋅ ( − 2 ) 1 ⋅ ( −
1 ) 1 ⋅ 1 2 ⋅ ( − 2 ) 2 ⋅ ( − 1 ) 2 ⋅ 1 3 ⋅ ( − 2 ) 3 ⋅ ( − 1 ) 3 ⋅ 1 ) = ( − 2 − 1
1 − 4 − 2 2 − 6 − 3 3 ) {\displaystyle {\begin{pmatrix}1\\2\\3\end{pmatrix}}\cdot
{\begin{pmatrix}-2\\-
1\\1\end{pmatrix}}^{T}={\begin{pmatrix}1\\2\\3\end{pmatrix}}\cdot {\begin{pmatrix}-
2&-1&1\end{pmatrix}}={\begin{pmatrix}1\cdot (-2)&1\cdot (-1)&1\cdot 1\\2\cdot (-
2)&2\cdot (-1)&2\cdot 1\\3\cdot (-2)&3\cdot (-1)&3\cdot
1\end{pmatrix}}={\begin{pmatrix}-2&-1&1\\-4&-2&2\\-6&-3&3\end{pmatrix}}}
{\displaystyle {\begin{pmatrix}1\\2\\3\end{pmatrix}}\cdot {\begin{pmatrix}-2\\-
1\\1\end{pmatrix}}^{T}={\begin{pmatrix}1\\2\\3\end{pmatrix}}\cdot {\begin{pmatrix}-
2&-1&1\end{pmatrix}}={\begin{pmatrix}1\cdot (-2)&1\cdot (-1)&1\cdot 1\\2\cdot (-
2)&2\cdot (-1)&2\cdot 1\\3\cdot (-2)&3\cdot (-1)&3\cdot
1\end{pmatrix}}={\begin{pmatrix}-2&-1&1\\-4&-2&2\\-6&-3&3\end{pmatrix}}}
⟨ A , B ⟩ = spur ( A T B ) = ∑ i = 1 n ∑ j = 1 m a i j b i j {\displaystyle
\left\langle A,B\right\rangle =\operatorname {spur} (A^{T}B)=\sum _{i=1}^{n}\sum
_{j=1}^{m}a_{ij}b_{ij}} \left\langle A,B\right\rangle = \operatorname{spur}(A^TB)
=\sum_{i=1}^n\sum_{j=1}^m a_{ij}b_{ij}
⟨ A , B ⟩ = spur ( A T ¯ B ) = ∑ i = 1 n ∑ j = 1 m a i j ¯ b i j
{\displaystyle \left\langle A,B\right\rangle =\operatorname {spur} ({\overline
{A^{T}}}B)=\sum _{i=1}^{n}\sum _{j=1}^{m}{\overline {a_{ij}}}b_{ij}} \left\langle
A,B\right\rangle = \operatorname{spur}(\overline{A^T}B)
=\sum_{i=1}^n\sum_{j=1}^m \overline{a_{ij}}b_{ij}
ein komplexes Skalarprodukt und der Matrizenraum wird zu einem unitären Vektorraum.
Dieses Skalarprodukt wird auch Frobenius-Skalarprodukt genannt. Die von dem
Frobenius-Skalarprodukt induzierte Norm heißt Frobeniusnorm und mit ihr wird der
Matrizenraum zu einem Banachraum.
Anwendungen
Zusammenhang mit linearen Abbildungen
Ist K {\displaystyle K} K sogar ein Körper, kann man statt der Spaltenvektorräume
beliebige endlichdimensionale K {\displaystyle K} K-Vektorräume V {\displaystyle V}
V und W {\displaystyle W} W (der Dimension n {\displaystyle n} n bzw. m
{\displaystyle m} m) betrachten. (Falls K {\displaystyle K} K ein kommutativer Ring
mit 1 ist, dann kann man analog freie K-Moduln betrachten.) Diese sind nach Wahl
von Basen v = ( v 1 , … , v n ) {\displaystyle v=(v_{1},\dotsc ,v_{n})}
{\displaystyle v=(v_{1},\dotsc ,v_{n})} von V {\displaystyle V} V und w = ( w 1 , …
, w m ) {\displaystyle w=(w_{1},\dotsc ,w_{m})} {\displaystyle w=(w_{1},\dotsc
,w_{m})} von W {\displaystyle W} W zu den Koordinatenräumen K n {\displaystyle
K^{n}} K^{n} bzw. K m {\displaystyle K^{m}} K^m isomorph, weil zu einem beliebigen
Vektor u ∈ V {\displaystyle u\in V} u\in V eine eindeutige Zerlegung in
Basisvektoren
v u = ( α 1 ⋮ α n ) ∈ K n {\displaystyle {}_{v}u={\begin{pmatrix}\alpha
_{1}\\\vdots \\\alpha _{n}\end{pmatrix}}\in K^{n}} {}_vu=\begin{pmatrix}\alpha_1 \\
\vdots \\ \alpha_n\end{pmatrix} \in K^n
f ( v j ) = ∑ i = 1 m a i j w i {\displaystyle f(v_{j})=\sum
_{i=1}^{m}a_{ij}w_{i}} f(v_j) = \sum_{i=1}^m a_{ij}w_i
mit Koordinatenvektor
w f ( v j ) = ( a 1 j ⋮ a m j ) ∈ K m . {\displaystyle
{}_{w}f(v_{j})={\begin{pmatrix}a_{1j}\\\vdots \\a_{mj}\end{pmatrix}}\in K^{m}.}
{}_wf(v_j)=\begin{pmatrix}a_{1j} \\ \vdots \\ a_{mj}\end{pmatrix} \in K^m.
Die Abbildung ist dann vollständig festgelegt durch die sog. Abbildungsmatrix
w f v = ( a 11 … a 1 n ⋮ ⋮ a m 1 … a m n ) ∈ K m × n , {\displaystyle
{}_{w}f_{v}={\begin{pmatrix}a_{11}&\ldots &a_{1n}\\\vdots &&\vdots \\a_{m1}&\ldots
&a_{mn}\end{pmatrix}}\in K^{m\times n},}
{}_{w}f_{v}={\begin{pmatrix}a_{{11}}&\ldots &a_{{1n}}\\\vdots &&\vdots
\\a_{{m1}}&\ldots &a_{{mn}}\end{pmatrix}}\in K^{{m\times n}},
denn für das Bild des o. g. Vektors u {\displaystyle u} u gilt
f ( u ) = ∑ i = 1 m ∑ j = 1 n a i j α j w i , {\displaystyle f(u)=\sum
_{i=1}^{m}\sum _{j=1}^{n}a_{ij}\alpha _{j}w_{i},} f(u) = \sum_{i=1}^m \sum_{j=1}^n
a_{ij}\alpha_j w_i,
x ( g ∘ f ) v = x g w ⋅ w f v {\displaystyle {}_{x}(g\circ
f)_{v}={}_{x}g_{w}\cdot {}_{w}f_{v}} {}_x(g\circ f)_v = {}_xg_w \cdot {}_wf_v
Somit ist die Menge der linearen Abbildungen von V {\displaystyle V} V nach W
{\displaystyle W} W wieder isomorph zu K m × n . {\displaystyle K^{m\times n}.}
K^{m\times n}. Der Isomorphismus f ↦ w f v {\displaystyle f\mapsto {}_{w}f_{v}}
f \mapsto {}_wf_v hängt aber von den gewählten Basen v {\displaystyle v} v und w
{\displaystyle w} w ab und ist daher nicht kanonisch: Bei Wahl einer anderen Basis
v ′ {\displaystyle v'} v' für V {\displaystyle V} V bzw. w ′ {\displaystyle w'} w'
für W {\displaystyle W} W wird derselben linearen Abbildung nämlich eine andere
Matrix zugeordnet, die aus der alten durch Multiplikation von rechts bzw. links mit
einer nur von den beteiligten Basen abhängigen invertierbaren m × m {\displaystyle
m\times m} m \times m- bzw. n × n {\displaystyle n\times n} n\times n-Matrix (sog.
Basiswechselmatrix) entsteht. Das folgt durch zweimalige Anwendung der
Multiplikationsregel aus dem vorigen Absatz, nämlich
w ′ f v ′ = w ′ e w W ⋅ w f v ⋅ v e v ′ V {\displaystyle
{}_{w'}f_{v'}={}_{w'}e_{w}^{W}\cdot {}_{w}f_{v}\cdot {}_{v}e_{v'}^{V}}
{}_{w'}f_{v'} = {}_{w'}e^W_w \cdot {}_wf_v \cdot {}_ve^V_{v'}
Bleibt eine Eigenschaft von Matrizen unberührt von solchen Basiswechseln, so ist es
sinnvoll, diese Eigenschaft auch basisunabhängig der entsprechenden linearen
Abbildung zuzusprechen.
Im Zusammenhang mit Matrizen oft auftretende Begriffe sind der Rang und die
Determinante einer Matrix. Der Rang ist (falls K {\displaystyle K} K ein Körper
ist) im angeführten Sinne basisunabhängig, und man kann somit vom Rang auch bei
linearen Abbildungen sprechen. Die Determinante ist nur für quadratische Matrizen
definiert, die dem Fall V = W {\displaystyle V=W} V=W entsprechen; sie bleibt
unverändert, wenn derselbe Basiswechsel im Definitions- und Wertebereich
durchgeführt wird, wobei beide Basiswechselmatrizen zueinander invers sind:
v ′ f v ′ = ( v e v ′ V ) − 1 ⋅ v f v ⋅ v e v ′ V {\displaystyle
{}_{v'}f_{v'}=({}_{v}e_{v'}^{V})^{-1}\cdot {}_{v}f_{v}\cdot {}_{v}e_{v'}^{V}}
{\displaystyle {}_{v'}f_{v'}=({}_{v}e_{v'}^{V})^{-1}\cdot {}_{v}f_{v}\cdot
{}_{v}e_{v'}^{V}}
Spezielle Matrizen
Eigenschaften von Endomorphismen
Orthogonale Matrizen
Eine reelle Matrix A {\displaystyle A} A ist orthogonal, wenn die zugehörige
lineare Abbildung das Standardskalarprodukt erhält, das heißt, wenn
bzw.
erfüllt.
Diese Matrizen stellen Spiegelungen, Drehungen und Drehspiegelungen dar.
Unitäre Matrizen
Sie sind das komplexe Gegenstück zu den orthogonalen Matrizen. Eine komplexe
Matrix A {\displaystyle A} A ist unitär, wenn die zugehörige Transformation die
Normierung erhält, das heißt, wenn
Projektionsmatrizen
Eine Matrix ist eine Projektionsmatrix, falls
gilt, sie also idempotent ist, das heißt, die mehrfache Anwendung einer
Projektionsmatrix auf einen Vektor lässt das Resultat unverändert. Eine idempotente
Matrix hat keinen vollen Rang, es sei denn, sie ist die Einheitsmatrix. Geometrisch
entsprechen Projektionsmatrizen der Parallelprojektion entlang des Nullraumes der
Matrix. Steht der Nullraum senkrecht auf dem Bildraum, so erhält man eine
Orthogonalprojektion.
Beispiel: Es sei X {\displaystyle X} X eine ( m × n ) {\displaystyle (m\times
n)} (m\times n)-Matrix und damit selbst nicht invertierbar. Falls der Rang von X
{\displaystyle X} X gleich n {\displaystyle n} n ist, dann ist ( X T X )
{\displaystyle (X^{T}X)} {\displaystyle (X^{T}X)} invertierbar und die ( m × m )
{\displaystyle (m\times m)} (m\times m)-Matrix
A = X ( X T X ) − 1 X T {\displaystyle A=X\,(X^{T}X)^{-1}X^{T}} A = X \,
(X^TX)^{-1}X^T
Nilpotente Matrizen
Eine Matrix N {\displaystyle N} N heißt nilpotent, falls eine Potenz N k
{\displaystyle N^{k}} N^k (und damit auch jede höhere Potenz) die Nullmatrix
ergibt.
Symmetrische Matrizen
Eine Matrix A {\displaystyle A} A heißt symmetrisch, wenn sie gleich ihrer
transponierten Matrix ist:
A T = A {\displaystyle A^{T}=A} A^T = A
( 1 2 3 2 4 5 3 5 6 ) T = ( 1 2 3 2 4 5 3 5 6 ) {\displaystyle
{\begin{pmatrix}1&2&3\\2&4&5\\3&5&6\end{pmatrix}}^{T}={\begin{pmatrix}1&2&3\\2&4&5\
\3&5&6\end{pmatrix}}} \begin{pmatrix} 1 & 2 & 3 \\ 2 & 4 & 5 \\ 3 & 5 & 6
\end{pmatrix}^T = \begin{pmatrix} 1 & 2 & 3 \\ 2 & 4 & 5 \\ 3 & 5 & 6 \end{pmatrix}
Hermitesche Matrizen
Hermitesche Matrizen sind das komplexe Analogon der symmetrischen Matrizen. Sie
entsprechen den hermiteschen Sesquilinearformen und den selbstadjungierten
Endomorphismen.
Eine Matrix A ∈ C n × n {\displaystyle A\in \mathbb {C} ^{n\times n}}
A\in\mathbb C^{n\times n} ist hermitesch oder selbstadjungiert, wenn gilt:
Schiefsymmetrische Matrizen
Eine Matrix A {\displaystyle A} A heißt schiefsymmetrisch oder auch
antisymmetrisch, wenn gilt:
( 0 1 2 − 1 0 3 − 2 − 3 0 ) {\displaystyle {\begin{pmatrix}0&1&2\\-1&0&3\\-
2&-3&0\end{pmatrix}}} {\displaystyle {\begin{pmatrix}0&1&2\\-1&0&3\\-2&-
3&0\end{pmatrix}}}
Weitere Konstruktionen
Enthält eine Matrix komplexe Zahlen, erhält man die konjugierte Matrix, indem man
ihre Komponenten durch die konjugiert komplexen Elemente ersetzt. Die adjungierte
Matrix (auch hermitesch konjugierte Matrix) einer Matrix A {\displaystyle A} A wird
mit A ∗ {\displaystyle A^{*}} A^* bezeichnet und entspricht der transponierten
Matrix, bei der zusätzlich alle Elemente komplex konjugiert werden.
Eine Übergangs- oder stochastische Matrix ist eine Matrix, deren Einträge alle
zwischen 0 und 1 liegen und deren Zeilen bzw. Spaltensummen 1 ergeben. Sie dienen
in der Stochastik zur Charakterisierung zeitlich diskreter Markow-Ketten mit
endlichem Zustandsraum. Ein Spezialfall hiervon sind die doppelt-stochastischen
Matrizen.
Unendlichdimensionale Räume
Auch für unendlichdimensionale Vektorräume (sogar über Schiefkörpern) gilt, dass
jede lineare Abbildung f : U → V {\displaystyle f\colon U\to V} f\colon U\to V
eindeutig durch die Bilder f ( u ) {\displaystyle f(u)} f(u) der Elemente u
{\displaystyle u} u einer Basis B U ⊂ U {\displaystyle {\mathcal {B}}_{U}\subset U}
\mathcal{B}_U\subset U bestimmt ist und diese beliebig gewählt werden und zu einer
linearen Abbildung auf ganz U {\displaystyle U} U fortgesetzt werden können. Ist
nun B V {\displaystyle {\mathcal {B}}_{V}} \mathcal{B}_V eine Basis von V
{\displaystyle V} V, so lässt sich f ( u ) {\displaystyle f(u)} f(u) eindeutig als
(endliche) Linearkombination von Basisvektoren schreiben, d. h., es existieren
eindeutige Koeffizienten f ( u ) b ∈ K {\displaystyle f(u)_{b}\in K} f(u)_b\in K
für b ∈ B V {\displaystyle b\in {\mathcal {B}}_{V}} b\in\mathcal{B}_V, von denen
nur endlich viele von null verschieden sind, sodass f ( u ) = ∑ b ∈ B V f ( u ) b b
{\displaystyle f(u)=\sum _{b\in {\mathcal {B}}_{V}}f(u)_{b}b}
f(u)=\sum_{b\in\mathcal{B}_V}f(u)_b b. Dementsprechend lässt sich jede lineare
Abbildung als möglicherweise unendliche Matrix auffassen, wobei jedoch in jeder
Spalte ( B U {\displaystyle {\mathcal {B}}_{U}} \mathcal{B}_U „nummeriere“ die
Spalten und die Spalte zu u {\displaystyle u} u bestehe dann aus den von den
Elementen von B V {\displaystyle {\mathcal {B}}_{V}} \mathcal{B}_V nummerierten
Koordinaten f ( u ) b {\displaystyle f(u)_{b}} f(u)_b) nur endlich viele Einträge
von null verschieden sind, und umgekehrt. Die entsprechend definierte
Matrixmultiplikation entspricht wiederum der Komposition linearer Abbildungen.
Belege
Matrix
Navigationsmenü
Nicht angemeldet
Diskussionsseite
Beiträge
Benutzerkonto erstellen
Anmelden
Artikel
Diskussion
Lesen
Bearbeiten
Quelltext bearbeiten
Versionsgeschichte
Suche
Hauptseite
Themenportale
Zufälliger Artikel
Mitmachen
Artikel verbessern
Neuen Artikel anlegen
Autorenportal
Hilfe
Letzte Änderungen
Kontakt
Spenden
Werkzeuge
In anderen Projekten
Commons
Drucken/exportieren
Buch erstellen
Als PDF herunterladen
Druckversion
In anderen Sprachen
Dansk
English
Español
Français
Hrvatski
Italiano
Nederlands
Русский
Türkçe
Links bearbeiten
Der Text ist unter der Lizenz „Creative Commons Attribution/Share Alike“
verfügbar; Informationen zu den Urhebern und zum Lizenzstatus eingebundener
Mediendateien (etwa Bilder oder Videos) können im Regelfall durch Anklicken dieser
abgerufen werden. Möglicherweise unterliegen die Inhalte jeweils zusätzlichen
Bedingungen. Durch die Nutzung dieser Website erklären Sie sich mit den
Nutzungsbedingungen und der Datenschutzrichtlinie einverstanden.
Wikipedia® ist eine eingetragene Marke der Wikimedia Foundation Inc.
Datenschutz
Über Wikipedia
Impressum
Entwickler
Statistiken
Stellungnahme zu Cookies
Mobile Ansicht
Wikimedia Foundation
Powered by MediaWiki