ABSTRACT
In this chapter, the authors propose an alternative solution to the problem of adaptive regularization by adopting a new global cost measure. They address the adaptive regularization problem by proposing a novel image model, the adoption of which in turn necessitates the use of powerful optimization algorithms such as those typical in the field of evolutionary computation. Evolutionary programming belongs to the class of optimization algorithms known as evolutionary computational algorithms that mimic the process of natural evolution to search for an optimizer of a cost function. Evolutionary Programming shares many common features with evolutionary strategy in that the primary adaptation operations are also carried out in the space of phenotypes. Genetic algorithm is the most widely used among the three evolutionary computational algorithms. A Hierarchical Cluster Model is a hierarchical neural network that coordinates parallel, distributed subnetworks in a recursive fashion with its cluster structures closely matching the homogeneous regions of an image.
