"P(GA102/02/0132)" . "Nottingham" . . "Nottingham" . "Binary or Real? Real-coded Binary!" . "University of Nottingham" . "This paper presents a novel approach to improve the performance of genetic algorithms - called genetic algorithms with real-coded binary representation (GARB). The proposed algorithm is capable of maintaining the population diversity during the whole run which protects it from premature convergence. This is achieved by using a special encoding scheme with a high redundancy, which is supported by so-called gene-strength adaptation mechanism. The proposed approach was evaluated on various test problems, all of them considered to be hard for standard genetic algorithms. The results show the algorithm significantly outperforms standard genetic algorithm and achieves results competitive with other techniques on non-stationary problems."@en . "201 ; 206" . "Nen\u00ED k dispozici"@cs . "Nen\u00ED k dispozici"@cs . "[4F6A7F97503C]" . "Binary or Real? Real-coded Binary!"@en . . . . "21230" . . "1"^^ . . "1-84233-110-8" . "Binary or Real? Real-coded Binary!"@en . . . "6"^^ . . "Nen\u00ED k dispozici"@cs . "1"^^ . "RIV/68407700:21230/04:03104112" . . "Kubal\u00EDk, Ji\u0159\u00ED" . . "genetic algorithms; premature convergence; redundancy; representation"@en . "RIV/68407700:21230/04:03104112!RIV/2005/GA0/212305/N" . . . . "2004-12-16+01:00"^^ . "Recent Advances in Soft Computing 2004" . . . . "Binary or Real? Real-coded Binary!" . "556144" . "This paper presents a novel approach to improve the performance of genetic algorithms - called genetic algorithms with real-coded binary representation (GARB). The proposed algorithm is capable of maintaining the population diversity during the whole run which protects it from premature convergence. This is achieved by using a special encoding scheme with a high redundancy, which is supported by so-called gene-strength adaptation mechanism. The proposed approach was evaluated on various test problems, all of them considered to be hard for standard genetic algorithms. The results show the algorithm significantly outperforms standard genetic algorithm and achieves results competitive with other techniques on non-stationary problems." .