diff --git a/slides-pdf/slides-classification-linear.pdf b/slides-pdf/slides-classification-linear.pdf
index 90bc0d6f5..3f4c425b1 100644
Binary files a/slides-pdf/slides-classification-linear.pdf and b/slides-pdf/slides-classification-linear.pdf differ
diff --git a/slides/supervised-classification/slides-classification-linear.tex b/slides/supervised-classification/slides-classification-linear.tex
index 63b0df60a..ec02c8d08 100644
--- a/slides/supervised-classification/slides-classification-linear.tex
+++ b/slides/supervised-classification/slides-classification-linear.tex
@@ -42,7 +42,7 @@
 
 \begin{itemize}
 \item $\bm{w}_k$ and $b_k$ do not necessarily refer to parameters $\thetav_k$, although they often coincide; discriminant simply must be writable in an affine-linear way 
-\item reasonsfor the transformation is that we only care about the position of the decision boundary
+\item reasons for the transformation is that we only care about the position of the decision boundary
 \end{itemize}