Cellular automata evolution using convolutional technique.
We utilize the new techniques deveveloped for neuronal networks, in particular, convolutional neuronal networks. The convolution method is used to obtain the new value of each cell generation to generation, the filter or kernel is multiplied with the neighborhood matrix of each cell, then a real function is evaluated in the product matrix so we can get a real value for the cell.
Future work will consist in the training of kernels as convolutional kernels do, so we can obtain, under a arbitrary function, the filter that gives a desire behavior in a cellular automata.