mirror of
https://gitlab.cs.fau.de/ik15ydit/latexandmore.git
synced 2024-12-26 01:56:05 +01:00
Komposition von Zufallsvektor
This commit is contained in:
parent
e8f92cc33e
commit
d43113ea90
@ -1,5 +1,16 @@
|
||||
\documentclass{article}
|
||||
\usepackage{amsmath}
|
||||
% -------- Umlaute korrekt ----------------
|
||||
\usepackage[utf8]{inputenc}
|
||||
\usepackage[ngerman, english]{babel}
|
||||
%-------------------------------------------
|
||||
|
||||
% TikZ Library
|
||||
\usepackage{tikz}
|
||||
\usetikzlibrary{arrows,backgrounds,positioning,fit,calc,petri}
|
||||
\usetikzlibrary{shapes, shapes.misc}
|
||||
\usetikzlibrary{decorations.markings,decorations.pathmorphing}
|
||||
% -----------------------------------------
|
||||
\DeclareMathSizes{10}{10}{10}{10}
|
||||
\title{Mathe C4 Merz - Cheatsheet}
|
||||
\author{Yannik Schmidt (Sheppy)\\September 2015}
|
||||
@ -234,9 +245,9 @@
|
||||
% \\ \\ \textbf{Moeglichkeit b) - Nach $x_1$ oder $x_2$ umstellen} \\ (ggf. mit Koordinatentransformation)
|
||||
\subsubsection{Alternatives Beispiel:}
|
||||
X,Y stochastisch unabhaengige, mit Parameter 'p' geometrisch verteilte Zufallsvarriablen in Wahrscheinlichkeitsraum $(\varOmega , \mathcal{A},P)$. Welche Verteilung besitzt
|
||||
Zufallsvarriable $Z = min(X,Y)$, definiert durch $Z( \omega ) = min \{X(\omega),Y(\omega)\}$.\\ \[
|
||||
|
||||
\]
|
||||
Zufallsvarriable $Z = min(X,Y)$, definiert durch $Z( \omega ) = min \{X(\omega),Y(\omega)\}$.\\
|
||||
|
||||
|
||||
\section{Marginaldichte - Beispielrechnung}
|
||||
\[
|
||||
f(x_z,x_2)=
|
||||
@ -264,23 +275,42 @@ und:
|
||||
\int \int f(x_1,x_2) dx_1 dx_2 = 1
|
||||
\]
|
||||
\section{Koordinatentransformation}
|
||||
\end{document}
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
|
||||
\section{Komposition von Zufallsvektoren}
|
||||
\begin{center}
|
||||
\begin{tikzpicture}[
|
||||
bend angle=45,
|
||||
scale = 1.5,
|
||||
pre/.style={<-,shorten <=1pt,>=stealth',semithick},
|
||||
post/.style={->,shorten >=1pt,>=stealth',semithick},
|
||||
mid/.style={-,shorten >=1pt,>=stealth',semithick},
|
||||
place/.style={circle,draw=red!50,fill=red!20,thick}]
|
||||
|
||||
\node[place] (A) at ( 0,0)[label=above:Before] {$(\Omega, A, P) $};
|
||||
\node[place] (B) at ( 2,0) {$(R^n, B_n, P^X)$}
|
||||
edge [pre] node [auto] {X} (A);
|
||||
\node[place, align=center] (C) at ( 2,-3) {$(R^m, B_m, P^G$}
|
||||
edge [pre] node [auto] {$Y = G \circ X$} (A)
|
||||
edge [pre] node [auto] {$G$} (B);
|
||||
\end{tikzpicture}
|
||||
\end{center}
|
||||
\vspace*{7pt}
|
||||
Problem: Verteilung von $P^G$ gesucht bei gegebenen $P^X$:
|
||||
\begin{enumerate}
|
||||
\item Funktionaldeterminante ($J_{G(x)}$) von $G$ berechnen
|
||||
\item Umkehrabbildung $G^*$ berechnen. Alle Zufallsvariablen werden
|
||||
werden mittels Funktionen verändert: z.B: $y_1 = x_1/x_2$.
|
||||
Jede i-te Funktion nach $x_i$ auflösen.
|
||||
\item Gesuchte Funktion: $g(y) = f(G^*(y))\frac{1}{|J_G(G^*(y))|}$\\
|
||||
$\longrightarrow$ Setze für alle $x_i$ dementsprechend $y_i$ ein und multipliziere
|
||||
mit Kehrwehrt von Funktionaldeterminante.
|
||||
\end{enumerate}
|
||||
|
||||
\begin{align}
|
||||
J_{G(x)} =
|
||||
\begin{pmatrix}
|
||||
\frac{\partial G_1}{\partial x_1} (x) & \cdots & \frac{\partial G_1}{\partial x_n} (x) \\
|
||||
\vdots & \ddots & \vdots \\
|
||||
\frac{\partial G_n}{\partial x_1} (x) & \cdots & \frac{\partial G_n}{\partial x_n} (x) \\
|
||||
\end{pmatrix}
|
||||
\end{align}
|
||||
\end{document}
|
||||
|
Loading…
Reference in New Issue
Block a user