Applying Advanced Learning Algorithms to ALVINN

ALVINN (Autonomous Land Vehicle in a Neural Net) is a Backpropagation trained neural network which is capable of autonomously steering a vehicle in road and highway environments. Although ALVINN is fairly robust, one of the problems with it has been the time it takes to train. As the vehicle is capable of on-line learning, the driver has to drive the car for about 2 minutes before the network is capable of autonomous operation. One reason for this is the use of Backprop. In this report, we describe the original ALVINN system, and then look at three alternative training methods Quickprop, Cascade Correlation, and Cascade 2. We then run a series of trials using Quickprop, Cascade Correlation and Cascade2, and compare them to a BackProp baseline. Finally, a hidden unit analysis is performed to determine what the network is learning. Applying Advanced Learning Algorithms to ALVINN Parag H. Batavia Dean A. Pomerleau Charles E. Thorpe 9 October 1996 CMU-RI-TR-96-31 This research was partially supported under a National Science Foundation Graduate Fellowship, the National Highway Traffic Safety Administration (NHTSA) under contract no. DTNH22-93-C-07023, and by USDOT under Cooperative Agreement Number DTFH61-94-X-00001 as part of the National Automated Highway System Consortium