The best currently known class of dynamically equivalent cellular automata rules for density classification

Non-uniform cellular automata based associative memory: Evolutionary design and basins of attraction

“…The genetic algorithm (GA) [11,12] provides a useful framework to arrive at the desired CA rule structures appropriate to model a physical system. The GA has been successfully applied to search for the desired CAs performing a specific task such as density classification [4,6,18,26], sequence generation [28], cryptography [31], modeling recrystallization process [7,23], pattern recognition [9,10], and so on.…”

Cellular Automata for Pattern Recognition

“…The cyclic state contains a pivotal point which is a solution to classification problem, while the transient states (all possible inputs) are contained in the non-cyclic states. The attractor cycle lengths (height) in the GMACA (Oliveira, et al, 2006;Sipper, 1996) are greater than or equal to one, while Multiple Attractor Cellular Automata (MACA) (Das, et al, 2008;Sipper, 1996) is limited to one. In Fig.…”

A Parallel Multiobjective Approach to Evolving Cellular Automata Rules by Cell State Change Dynamics

“…Oliveira et al [14] present a multiobjective evolutionary approach to DCT based on the nondominated sorting genetic algorithm (NSGA). The algorithm is implemented in a distributed cooperative computational environment and was able to discover new rules with a slightly better performance of 0.8616.…”

“…An important constraint included in the latter two approaches [14], [28] presented above refers to the number of state transitions of a rule following black-white, leftright (BWLR) transformations. Experiments have shown that this BWLR symmetry should be maximal to achieve highperformant DCT rules [25].…”

“…Parallel computers are architectures that efficiently support the approach of evolving through computational simulation as the majority of these simulations are independent and can be executed on different computational units. Thus, the impediment regarding the computational costs of evolving CA can be alleviated by parallelization [10], [11], [12], [13], [14], [15], [16]. The integration of the multiobjective EA with the proposed GPU parallelization scheme results in better memory latency hiding and scaling compared to recent GPU schemes for evolutionary CA design [17] which use thousand or at most ten of thousands threads.…”