Expanded bootstrap chapter and improved latex header

This commit is contained in:
2015-11-01 22:39:06 +01:00
parent 40a5343fa8
commit 5edac6cfd6
15 changed files with 419 additions and 65 deletions

View File

@@ -0,0 +1,3 @@
>> mlegammafit
shape=2.10
scale=0.95

View File

@@ -0,0 +1,5 @@
>> mlemean
standard deviation of the data is 1.93
mean of the data is 2.98
maximum of likelihood is at 2.98
maximum of log-likelihood is at 2.98

View File

@@ -105,7 +105,7 @@ x$ der Daten. D.h. das arithmetische Mittel maximiert die
Wahrscheinlichkeit, dass die Daten aus einer Normalverteilung mit
diesem Mittelwert gezogen worden sind (\figref{mlemeanfig}).
\begin{exercise}[mlemean.m]
\begin{exercise}{mlemean.m}{mlemean.out}
Ziehe $n=50$ normalverteilte Zufallsvariablen mit einem Mittelwert $\ne 0$
und einer Standardabweichung $\ne 1$.
@@ -244,7 +244,7 @@ gesuchten Wahrscheinlichkeitsdichtefunktion bei der die Log-Likelihood
nichtlinieares Optimierungsproblem, das mit numerischen Verfahren, wie
z.B. dem Gradientenabstieg, gel\"ost wird \matlabfun{mle}.
\begin{exercise}[mlegammafit.m]
\begin{exercise}{mlegammafit.m}{mlegammafit.out}
Erzeuge Gammaverteilte Zufallszahlen und benutze Maximum-Likelihood,
um die Parameter der Gammafunktion aus den Daten zu bestimmen.
\newpage

View File

@@ -94,7 +94,7 @@ ax.annotate('',
xytext=(maxp, -30), textcoords='data',
arrowprops=dict(arrowstyle="->", relpos=(0.5,0.5),
connectionstyle="angle3,angleA=80,angleB=90") )
ax.text(maxp+0.05, -1300, 'most\nlikely\norientation')
ax.text(maxp+0.05, -1100, 'most likely\norientation\ngiven the responses')
ax.plot(phases, loglikelihood, '-b')
plt.savefig('mlecoding.pdf')