Abstract
The concepts are presented of a neural model based shell that integrates artificial neural networks (ANN) and artificial intelligence (AI) for problem solving. The shell may possess inherited and learnt ANN and AI subsystems. The shell has and develops (i) cues to the environment for dimensionality reduction, (ii) rules between elements of the reduced dimensional internal representation, (iii) `concepts' for achieving goals, i.e. for solving existing problems, (iv) the shell then causes the concepts to compete in order to come to a decision. The shell is designed for control problems, e.g. robotic tasks, control of plants, investment advisory systems, and may have very different ANN and AI parts. Here, we consider a simple robotic-like object in two dimensional space.
Users
Please
log in to take part in the discussion (add own reviews or comments).