From b6aa0b124de02bda659db226ef66f1886f98252e Mon Sep 17 00:00:00 2001 From: Stratis Ioannidis Date: Thu, 1 Nov 2012 23:42:57 -0700 Subject: stratis comments --- problem.tex | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) (limited to 'problem.tex') diff --git a/problem.tex b/problem.tex index 468ab01..03ce49b 100644 --- a/problem.tex +++ b/problem.tex @@ -47,9 +47,9 @@ incentivize her participation in the study. A mechanism $\mathcal{M} = (f,p)$ comprises (a) an \emph{allocation function} $f:\reals_+^n \to 2^\mathcal{N}$ and (b) a \emph{payment function} -$p:\reals_+^n\to \reals_+^n$. The allocation function determines, given the -vector or costs $[c_i]_{i\in\mathcal{N}}$, the set -$S\subseteq \mathcal{N}$ of experiments to be conducted. The payment function +$p:\reals_+^n\to \reals_+^n$. Given the +vector or costs $[c_i]_{i\in\mathcal{N}}$, the allocation function determines the set +$S\subseteq \mathcal{N}$ of experiments to be conducted, while the payment function returns a vector of payments $[p_i]_{i\in \mathcal{N}}$. As in \cite{singer-mechanisms, chen}, we study mechanisms that are normalized ($i\notin S$ implies $p_i=0$), individually rational ($p_i\geq c_i$) and have @@ -102,6 +102,8 @@ $\inf\{c_i: i\in f(c_i, c_{-i})\}$. \end{itemize} \end{theorem} +\stratis{Explain why this is important and what it implies about the things we need to prove. Also, don't overuse bullets.} + \begin{comment} \subsection{Experimental Design} -- cgit v1.2.3-70-g09d2