Extracting Refined Rules from Hybrid Neuro-Symbolic Systems

In this work is presented the preliminary studies of the development of a system for the extraction of refined rules in artificial neuronal networks from hybrid neuro-symbolic systems. The artificial neuronal networks (ANN) is a distributed massively parallel processor that is prone for the nature to store experimental knowledge and to make it available for the use. However a disadvantage of the ANN is they are considered "black boxes", since they transform the entrances in exits without noticing as this transformation is made.