Evolutionary Growth Perceptrons

The combination of genetic algorithms and parallel processing can be very powerful when applied to the design and learning of Multilayer Perceptron neural networks. It is well known that the definition of Multilayer Perceptron neural networks is a very difficult task that could become, to a large extent, a process of trial and error. Also, the learning process of Multilayer Perceptron neural networks can be very slow, putting in danger the performance of countless applications. Therefore techniques for automating the design of neural networks are clearly of interest and the use of parallel processing is essential in minimizing the time required on the learning process. Furthermore, the use of cooperation in the genetic algorithm allows the interaction of different populations, avoiding local minima and helping in the search of the ideal solution. Finally, the use of exclusive evolution behavior on each copy of the genetic algorithm running in each task helped in enhancing the diversity of populations and the search for the fit solution.