blob: 121788a65521e6f39a8dd90fac8d62a4d804a00a (
plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
|
\begin{frame}
\frametitle{Positionnement}
\centering
\input{images/tikz/ckoi/pos22}
\end{frame}
{
\usebackgroundtemplate{\includegraphics[width=\paperwidth]{images/background/card/background.pdf}}
\begin{frame}
%\vspace{70px}
\hspace{70px}
\begin{minipage}{250px}
\centering
\Large
\textcolor{accent}{
Comment protéger l'attribut sensible contre l'AIA ?
}
\normalsize
\vspace{20px}
Lien entre équité et AIA
\end{minipage}
\end{frame}
}
\begin{frame}
\frametitle{Définitions de l'équité de la protection contre l'AIA}
\begin{definition}[Parité démographique]
$f$ satisfait la parité démographique pour $S$ si et seulement si
\begin{equation*}
P_{(f\circ X,S)}=
P_{f\circ X}\otimes P_S
\end{equation*}
\end{definition}
\pause
\begin{definition}[S est protégé]
$S$ est protégé contre l'AIA sur $f$ si et seulement si
\begin{equation*}
\forall a\in A~
P_{(a\circ f\circ X,S)}=
P_{a\circ f\circ X}\otimes P_S
\end{equation*}
où $A$ est l'ensemble des fonctions mesurables de $F$ dans $G$.
\end{definition}
\end{frame}
\begin{frame}
\frametitle{Lien entre l'équité, la confidentialité et l'exactitude équilibrée}
\begin{theorem}
Les trois propositions suivantes sont équivalentes.
\begin{itemize}
\item[$(\alpha)$] $f$ satisfait la parité démographique pour $S$
\item[$(\beta)$] $S$ est protégé contre l'AIA sur $f$
\item[$(\gamma)$] $\max_{a\in A} BA(f) = \frac{1}{\#G}$
\end{itemize}
\end{theorem}
\end{frame}
\begin{frame}
\frametitle{Eléments de démonstration}
$(\alpha) \iff (\beta)$ grâce au Lemme 4.1 de mon manuscrit de thèse.
\pause
\begin{align*}
&(\beta)&\\
\iff&\forall a\in A~P_{(a\circ f\circ X,S)} =
P_{a\circ f\circ X}\otimes P_S&\text{Définition de la protection.}\\
\iff&\forall a\in A~BA(a) = \frac{1}{\#G}&\text{Théorème 4.2 du manuscrit.}\\
\iff&\text{max}_{a\in A}BA(a) = \frac{1}{\#G}&\text{Conséquence du Théorème 5.2 du manuscrit.}\\
\iff&(\gamma)&
\end{align*}
\pause
Nous avons donc bien $(\alpha)\iff(\beta)\iff(\gamma)$.
\end{frame}
\begin{frame}
\frametitle{Cas particulier : classification binaire et attribut binaire}
\begin{definition}[Niveau de parité démographique]
\begin{equation*}
DemParLvl = |P(f\circ X=0\mid S=0) -
P(f\circ X=0\mid S=1)|
\end{equation*}
\end{definition}
\pause
\begin{theorem}
\begin{equation*}
\text{max}_{a\in A}BA(a) =
\frac{1}{2}(1+DemParLvl)
\end{equation*}
\end{theorem}
\end{frame}
|