Issues and methods for learning by autonomous robots
Brian J. Tillotson, D.L. Johnson · 1989
New issues arise when machine learning techniques are applied to real autonomous robots. New issues described in this dissertation are example bounding, learning from sensor data, topic selection, and problem solving while learning. This dissertation presents methods which address these issues and which are responsive to a robot's need for computational efficiency and robustness. The methods were implemented in two computer programs which were tested using real and simulated robots. The achievements of the programs are discussed.