1 files changed, 12 insertions, 9 deletions

diff --git a/paper/sections/intro.tex b/paper/sections/intro.tex
index 264476b..f369e3c 100644
--- a/paper/sections/intro.tex
+++ b/paper/sections/intro.tex
@@ -117,10 +117,10 @@ achieves a ${\cal O}(s \log m)$ guarantee in the case of tree graphs. The work
 of~\cite{Abrahao:13} studies the same continuous-model framework as
 \cite{GomezRodriguez:2010} and obtains an ${\cal O}(s^9 \log^2 s \log m)$
 support recovery algorithm, without the \emph{correlation decay} assumption.
+\cite{du2013uncover} propose a similar algorithm to ours for recovering the
+weights of the graph under a continuous-time independent cascade model, without
+proving theoretical guarantees.
 
-{\color{red} Du et.~al make a citation}
-
-{\color{red} say they follow the same model as Gomez and abrahao}
 Closest to this work is a recent paper by \citet{Daneshmand:2014}, wherein the
 authors consider a $\ell_1$-regularized objective function. They adapt standard
 results from sparse recovery to obtain a recovery bound of ${\cal O}(s^3 \log
@@ -132,12 +132,15 @@ Independent Cascade model under weaker assumptions. Furthermore, we analyze both
 the recovery of the graph's edges and the estimation of the model's parameters,
 and achieve close to optimal bounds.
 
-\begin{comment}
-Their work has the merit of studying a generalization of the discrete-time
-independent cascade model to continuous functions. Similarly to
-\cite{Abrahao:13}, they place themselves in the restrictive single-source
-context.
-\end{comment}
+The work of~\cite{du2014influence} is slightly orthogonal to ours since they
+suggest learning the \emph{influence} function, rather than the networks
+parameters directly.
+%\begin{comment}
+%Their work has the merit of studying a generalization of the discrete-time
+%independent cascade model to continuous functions. Similarly to
+%\cite{Abrahao:13}, they place themselves in the restrictive single-source
+%context.
+%\end{comment}
 
 \begin{comment}
 \paragraph{Our contributions}