These devices sometimes have some limitations in their design and operation when intending to provide a solution to a particular problem.However, these systems are still in a transformation process; that is, the development of the networks is still open to changes.In other words, new paradigms must be developed, such as learning and the massive parallelism of the neural networks.