The goal of this research project is to develop an efficient scalable evolutionary algorithm for search, optimization, and machine learning by following the underlying mechanism of the natural process of gene expression (DNA -> mRNA -> Protein). This work also applies the developed algorithms to solve a distributed data mining problem. The approach consists of: (1) abstracting the transcription (DNA -> mRNA), translation (mRNA -> Protein sequence), and folding (sequence to 3D folded structure of protein) operations in the light of probabilistic and approximate representation construction; (2) development of search operators using the developed understanding; and (3) incorporation of these mechanisms into an experimental distributed data mining system for discovering knowledge from distributed data. A successful completion this work will greatly enhance the field of evolutionary adaptive algorithms and its applications. It will offer a scalable approach for solving large optimization, machine learning, and data mining problems. Moreover, it will offer a fundamental understanding of the natural gene expression process from the perspective of evolutionary computation. www.eecs.wsu.edu/~hillol
