Luigi Stringa

Learn More
An approximation of the Bayes decision rule and its implementation on a two-layered network are described. The net is trained in two phases: first, probabilities of the discrete-valued input features are learnt by applying a Good-Turing based estimator; second, net weights are estimated by applying an adaptive gradient descent technique. Experiments were(More)
A sonar array and a camera are the sensors of a robot that learns to move inside a building. Learning is based on the association of ultrasonic and visual images with motor commands. The two sensors extract diierent information from the environment. Sonars produce a short-range depth image on the sonar array plane, vision gives a wide-range at image of the(More)