Scalability of a Hybrid Extended Compact Genetic Algorithm for Ground State Optimization of Clusters Export

@article{citeulike:1433755, abstract = {We analyze the utility and scalability of extended compact genetic algorithm (eCGA)\&mdash;a genetic algorithm (GA) that automatically and adaptively mines the regularities of the fitness landscape using machine learning methods and information theoretic measures\&mdash;for ground state optimization of clusters. In order to reduce the computational time requirements while retaining the high reliability of predicting near-optimal structures, we employ two efficiency-enhancement techniques: (1) hybridizing eCGA with a local search method, and (2) seeding the initial population with lowest energy structures of a smaller cluster. The proposed method is exemplified by optimizing silicon clusters with 4\&ndash;20 atoms. The results indicate that the population size required to obtain near-optimal solutions with 98\% probability scales sub linearly (as \&b.Theta;(<i>n</i>^0.83)) with the cluster size. The total number of function evaluations (cluster energy calculations) scales sub-cubically (as \&b.Theta;(<i>n</i>^2.45)), which is a significant improvement over exponential scaling of poorly designed evolutionary algorithms.}, author = {Sastry, Kumara and Goldberg, David and Johnson, D. D.}, citeulike-article-id = {1433755}, citeulike-linkout-0 = {http://dx.doi.org/10.1080/10426910701319654}, doi = {10.1080/10426910701319654}, journal = {Materials and Manufacturing Processes}, keywords = {ecga, edas, genetic-algorithms}, number = {5}, pages = {570--576}, posted-at = {2007-07-04 20:18:35}, priority = {2}, publisher = {Taylor \&amp; Francis}, title = {Scalability of a Hybrid Extended Compact Genetic Algorithm for Ground State Optimization of Clusters}, url = {http://dx.doi.org/10.1080/10426910701319654}, volume = {22}, year = {2007} }

TY - JOUR ID - citeulike:1433755 L3 - citeulike-article-id:1433755 N2 - We analyze the utility and scalability of extended compact genetic algorithm (eCGA)&mdash;a genetic algorithm (GA) that automatically and adaptively mines the regularities of the fitness landscape using machine learning methods and information theoretic measures&mdash;for ground state optimization of clusters. In order to reduce the computational time requirements while retaining the high reliability of predicting near-optimal structures, we employ two efficiency-enhancement techniques: (1) hybridizing eCGA with a local search method, and (2) seeding the initial population with lowest energy structures of a smaller cluster. The proposed method is exemplified by optimizing silicon clusters with 4&ndash;20 atoms. The results indicate that the population size required to obtain near-optimal solutions with 98% probability scales sub linearly (as &b.Theta;(<i>n</i>^0.83)) with the cluster size. The total number of function evaluations (cluster energy calculations) scales sub-cubically (as &b.Theta;(<i>n</i>^2.45)), which is a significant improvement over exponential scaling of poorly designed evolutionary algorithms. IS - 5 JF - Materials and Manufacturing Processes EP - 576 TI - Scalability of a Hybrid Extended Compact Genetic Algorithm for Ground State Optimization of Clusters PB - Taylor &amp; Francis VL - 22 SP - 570 KW - ecga KW - edas KW - genetic-algorithms AU - Sastry, Kumara AU - Goldberg, David AU - Johnson, DD PY - 2007/// UR - http://dx.doi.org/10.1080/10426910701319654 ER -