deletions | additions       diff --git a/problems.tex b/problems.tex  index 1b88337..57cebcd 100644  --- a/problems.tex  +++ b/problems.tex 
...
\section{Problems}  Using Given  the above setup, one can come up with setup as outlined above, there are  several interesting problems.   Some are very easy, others might be very complicated.  We list a few that problems one can investigate.  In our case,  we are interesting interested  in exploring, where we want comparing two approaches  to start simple and build more complicated algorithms as we go. learning to herd:  \begin{enumerate}  \item A dog has Dogs act independently, while they can observe each other, they cannot explicitly coordinate their movements.  \item There is a shepherd AI that coordinates the movements of the dogs. This is a two=tiered approach, where dogs are trained  to move perform specific smaller tasks (move  to a target \item A dog has location, circle the flock clockwise, etc.) and a central AI that gives the instructions  to move one sheep the dogs.  \end{enumerate}  To investigate these two approaches, we propose  to start with some simple tasks, and expand upon them.  Some examples of simple tasks include assigning  a target \item One for the  doghas  to move multiple sheep to a target  \item One too, and one  dog must move multiple sheep, each to a different target  \item Multiple dogs have having  to move multiple one  sheep to a target.  \item Learn dogs the target location.  More complicated tasks can be based upon these  basic herding commands\footnote{see \url{http://en.wikipedia.org/wiki/Herding_dog} under `Basic herding dog commands' for an overview} and then have a seperate AI give commands to the dogs.  \end{enumerate} tasks.