Results of Evolution Supervised by Genetic Algorithms

Authors

  • Lorentz JÄNTSCHI Technical University of Cluj-Napoca, Department of Chemistry, 103-105 Muncii Bvd., 400641 Cluj-Napoca (RO)
  • Sorana D. BOLBOACĂ Iuliu Hatieganu University of Medicine and Pharmacy, Department of Medical Informatics and Biostatistics, 400349 Cluj-Napoca (RO)
  • Mugur C. BĂLAN Technical University of Cluj-Napoca, Department of Chemistry, 103-105 Muncii Bvd., 400641 Cluj-Napoca (RO)
  • Radu E. SESTRAS University of Agricultural Sciences and Veterinary Medicine Cluj-Napoca, 3-5 Manastur, 400372 Cluj-Napoca (RO)
  • Mircea V. DIUDEA Babes-Bolyai University, Faculty of Chemistry and Chemical Engineering, Department of Organic Chemistry, 11 Arany Janos Str., 400028 Cluj-Napoca (RO)

DOI:

https://doi.org/10.15835/nsb234873

Keywords:

genetic algorithm (GA), evolution, genetic operators

Abstract

The efficiency of a genetic algorithm is frequently assessed using a series of operators of evolution like crossover operators, mutation operators or other dynamic parameters. The present paper aimed to review the main results of evolution supervised by genetic algorithms used to identify solutions to agricultural and horticultural hard problems and to discuss the results of using a genetic algorithms on structure-activity relationships in terms of behavior of evolution supervised by genetic algorithms. A genetic algorithm had been developed and implemented in order to identify the optimal solution in term of estimation power of a multiple linear regression approach for structure-activity relationships. Three survival and three selection strategies (proportional, deterministic and tournament) were investigated in order to identify the best survival-selection strategy able to lead to the model with higher estimation power. The Molecular Descriptors Family for structure characterization of a sample of 206 polychlorinated biphenyls with measured octanol-water partition coefficients was used as case study. Evolution using different selection and survival strategies proved to create populations of genotypes living in the evolution space with different diversity and variability. Under a series of criteria of comparisons these populations proved to be grouped and the groups were showed to be statistically different one to each other. The conclusions about genetic algorithm evolution according to a number of criteria were also highlighted.

Metrics

Metrics Loading ...

Downloads

Published

2010-09-27

How to Cite

JÄNTSCHI, L., BOLBOACĂ, S. D., BĂLAN, M. C., SESTRAS, R. E., & DIUDEA, M. V. (2010). Results of Evolution Supervised by Genetic Algorithms. Notulae Scientia Biologicae, 2(3), 12–15. https://doi.org/10.15835/nsb234873

Issue

Section

Review articles
CITATION
DOI: 10.15835/nsb234873

Most read articles by the same author(s)

1 2 3 4 5 6 > >>