Cooperative phenomena are observed in many complex dynamical systems in physics, biology, and even economics, and 1/f noise is considered to be a footprint of the complex dynamical systems in a critical state. Cooperations have been also studied in the field of evolutionary algorithms inspired by biological evolution, however, examination from the viewpoint of emergence of 1/f noise has not been done, and no cooperative architecture exhibiting 1/f noise has been proposed in the field. In order to study from such a viewpoint, a fusion of a simple genetic algorithm and a Boltzmann machine (FSGABM) is proposed as a new evolutionary algorithm. In order to compare the performance of the FSGABM with the simple genetic algorithm (SGA) which has no such cooperation, a test problem, i.e., a pattern restoration, and evaluation methods are also proposed. As a result of cooperation between the BM and the mutation operator, the FSGABM evolved toward the 'edge of evolution', which is a critical state, remaining in balance. Then, the FSGABM was able to generate effective evolution, even though the number of individuals in a generation was very small, and when the 1/f power spectrum in the low frequency region was examined, it implied moderate long-range time-correlation. In contrast, the SGA could not do so.

1/f noise, cooperative phenomena, evolutionary algorithm, genetic algorithm, Boltzmann machine.