From 90211ba1847251c552361a7d0b2eef1c540ef72a Mon Sep 17 00:00:00 2001
From: Thibaut Horel <thibaut.horel@gmail.com>
Date: Tue, 14 Apr 2015 15:27:24 -0400
Subject: Nips format and minor tweaks. We are already close to the page limit…
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

---
 paper/sections/negative.tex | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'paper/sections')

diff --git a/paper/sections/negative.tex b/paper/sections/negative.tex
index f88117a..2eaeb47 100644
--- a/paper/sections/negative.tex
+++ b/paper/sections/negative.tex
@@ -25,7 +25,7 @@ guarantees are sufficient for optimization. First look at learning weights in
 a cover function. Maybe facility location? Sums of concave over modular are
 probably too hard because of the connection to neural networks.
 
-\subsection{Optimizable $n\Rightarrow$ Learnable}
+\subsection{Optimizable $\nRightarrow$ Learnable}
 
 Recall the matroid construction from~\cite{balcan2011learning}:
 \begin{theorem}
-- 
cgit v1.2.3-70-g09d2