update images

Author: Accacio

code/example_introduction/gen_image.py

@@ -66,6 +66,6 @@
 plt.savefig(outputFolder + "/example_J" +  ".png",facecolor=fig.get_facecolor())
 
 
-ax.set_title('Fonctions objectives global and locales $J^\star$ and  $J_i^\star$', fontsize=16,usetex=True)
+ax.set_title('Fonctions objectif global et locales $J^\star$ and  $J_i^\star$', fontsize=16,usetex=True)
 plt.savefig(outputFolder + "/example_J_fr" +  ".pdf",facecolor=fig.get_facecolor())
 plt.savefig(outputFolder + "/example_J_fr" +  ".png",facecolor=fig.get_facecolor())

img/example_introduction/example_J.pdf

@@ -683,7 +683,6 @@ \section{Adapting detection and mitigation methods}\label{sec:detection-mitigati
 We propose to use a similar strategy to the one used in the last chapter.
 However, some difficulties arise.
 We do not have one tuple $(\Plin,\sik)$ for each subsystem, but we have $2^{\predhorz c}$ different tuples.
-
 Here follow proposals on how to counter the difficulties.
 
 \subsection{Mitigation and artificial scarcity}\label{sec:mitigation_ineq}
@@ -842,7 +841,7 @@ \subsection{Multiple Parameter Estimation}\label{sec:cons-about-mult}
 In the next subsection we describe the method more precisely.
 
 \subsubsection{Expectation Maximization}
-The main objective of the \EM{} algorithm is to find, from a set of observable data $\set{B}$, estimators of a set of parameters $\set{P}$ that maximize the log marginal likelihood of the observed data ${\ln\probability{\set{B};\set{P}}}$. The models generally have latent variables (unobservable) in a set $\set{U}$.
+The objective of the \EM{} algorithm is to find, from a set of observable data $\set{B}$, estimators of a set of parameters $\set{P}$ that maximize the log marginal likelihood of the observed data ${\ln\probability{\set{B};\set{P}}}$. The models generally have latent variables (unobservable) in a set $\set{U}$.
 
 The problem is that maximizing ${\ln\probability{\set{B};\set{P}}}$ does not have an analytical solution.
 So, the algorithm solves the optimization problem in a iterative manner.