Evolution of evolution: Self-constructing Evolutionary Turing Machine case study
Eugene Eberbach, Mark Burgin · 2007
The goal of this paper is to study the process of evolution of evolution. In other words, we study evolution that adapts evolutionary algorithms in parallel with their solutions. For this purpose, we define and investigate several extensions of evolutionary Turing machine model: self-constructing evolutionary Turing machines (SETM), self-constructing evolutionary Turing machines with a basic constructor (SBETM), self-constructing evolutionary Turing machines with a basic constructor and control information (CSBETM), and self-constructing evolutionary Turing machines with evolvable control information (CSETM). Such properties as expressiveness and complexity of different types of self-constructing evolutionary Turing machines are studied. It is demonstrated how self-constructive abilities allow one to essentially increase efficiency of evolutionary processes in general and evolutionary computations, in particular. We also investigate computation and construction universatility in the context of these new models.