Zastosowanie procesów epigenetycznych w algorytmach genetycznych

Abstract

The main purpose of the dissertation is to develop and verify the possibility of application mechanism based on epigenetics processes in genetics algorithms. In the dissertation three new operations in genetic algorithms were introduced. Epigenetics is the study of heritable phenotype changes that do not involve changes in the DNA sequence. The epigenetics processes was discovered later than the genetic algorithms was develop. Therefore, the epigenetic processes have not been included in classical genetic algorithms. In the dissertation the operations mimics the following epigenetic processes: prion inheritance, methylation of cytosine and allelic exclusion. The mentioned operations have been tested in terms of the impact on the efficiency of genetic algorithms, evaluating the reduction of the number of generations needed to obtain the expected result by the genetic algorithm, and the impact on the operation time of the algorithm. Research was carried out on four selected genetic algorithms: loading optimization, data grouping, outliers detection and biological sequence alignment. It was proved, based on the conducted research, that the application of each proposed operation in all algorithms used in the experiments influenced the reduction of the number of generations needed to obtain the expected result by algorithms. The following contents were included in the work: • description of optimization problems, • literature review on evolutionary algorithms, • the idea and operating principle of genetic algorithms were discussed, • literature review of existing modifications of genetic algorithms had been made, • discussed selected genetic algorithms, algorithms that were used as a algorithms of reference in the experiments, the results of which were included in the work, • the idea and method of implementation of proposed operations inspired by epigenetic processes been shown • the result of the experiments on the assessment of the impact of the proposed operations on the genetic algorithms was presented. Based on the results of the experiments presented in the work, the thesis of the dissertation has been proved.