Using this XML learning.xml over 6 hour period we collected this file dataBeforeProcessing.txt.
First the data was run through a few scripts to clean away X-Plane's graphical information, so only data collected by the plugin was left and so that the data would be MATLAB friendly.
Here is the end result data.txt.
Using LSPI (Least-Squares Policy Iteration) we produced approximate Q-functions, which define our policy.
Using this XML learned.xml that has the Q-function weights from LSPI the plane was able to take off.