## Further Experimentations On The Scalability Of The GEMGA (1998)

Venue: | In Lecture Notes in Computer Science: Parallel Problem Solving from Nature |

Citations: | 5 - 3 self |

### BibTeX

@INPROCEEDINGS{Kargupta98furtherexperimentations,

author = {Hillol Kargupta and Sanghamitra Bandyopadhyay},

title = {Further Experimentations On The Scalability Of The GEMGA},

booktitle = {In Lecture Notes in Computer Science: Parallel Problem Solving from Nature},

year = {1998},

pages = {315--324},

publisher = {Springer-Verlag}

}

### Years of Citing Articles

### OpenURL

### Abstract

. This paper reports the recent developments of the Gene Expression Messy Genetic Algorithm (GEMGA) research. It presents extensive experimental results for large problems with massive multi-modality, non-uniform scaling, and overlapping sub-problems. All the experimental results corroborate the linear time performance of the GEMGA for a wide range of problems, that can be decomposed into smaller overlapping and non-overlapping sub-problems in the chosen representation. These results further support the scalable performance of the GEMGA. 1 Introduction The recent past has witnessed a growing interest in designing scalable evolutionary algorithms that inductively detect the decomposable partitions (sometimes called genetic linkage in the Genetic Algorithm (GA) literature[8]) of the optimization problem. Messy genetic algorithms [4, 6, 5, 7, 10, 3], Dependency trees [2], distribution estimation [15, 16], are some examples. For problems that can be decomposed into smaller subproblems in ...

### Citations

3144 |
Adaptation in natural and artificial systems
- Holland
- 1992
(Show Context)
Citation Context ...itnessed a growing interest in designing scalable evolutionary algorithms that inductively detect the decomposable partitions (sometimes called genetic linkage in the Genetic Algorithm (GA) literature=-=[8]-=-) of the optimization problem. Messy genetic algorithms [4, 6, 5, 7, 10, 3], Dependency trees [2], distribution estimation [15, 16], are some examples. For problems that can be decomposed into smaller... |

344 |
Messy genetic algorithms: Motivation, analysis, and first results
- Goldberg, Korb, et al.
- 1989
(Show Context)
Citation Context ...ary algorithms that inductively detect the decomposable partitions (sometimes called genetic linkage in the Genetic Algorithm (GA) literature[8]) of the optimization problem. Messy genetic algorithms =-=[4, 6, 5, 7, 10, 3]-=-, Dependency trees [2], distribution estimation [15, 16], are some examples. For problems that can be decomposed into smaller subproblems in the chosen representation, these algorithms are likely to w... |

257 |
A Connectionist Machine for Genetic Hillclimbing
- Ackley
- 1987
(Show Context)
Citation Context ...f the functions is constructed by concatenating order5 subfunctions (both overlapping and non-overlapping versions are considered). These subfunctions are described below. i) Trap: The deceptive trap =-=[1]-=- function is defined as, f(x) = k if u = k; f(x) = k \Gamma 1 \Gamma u otherwise; where u is the unitation variable, or the number of 1-s in the string x, and k is the length of the subfunction. If we... |

255 | From recombination of genes to the estimation of distributions I. Binary parameters
- Mühlenbein, Paaß
- 1996
(Show Context)
Citation Context ...(sometimes called genetic linkage in the Genetic Algorithm (GA) literature[8]) of the optimization problem. Messy genetic algorithms [4, 6, 5, 7, 10, 3], Dependency trees [2], distribution estimation =-=[15, 16]-=-, are some examples. For problems that can be decomposed into smaller subproblems in the chosen representation, these algorithms are likely to work effectively by reducing the global search problem in... |

124 | Davies: Using optimal dependency-trees for combinatorial optimization: learning the structure of the search space
- Baluja
- 1997
(Show Context)
Citation Context ... the decomposable partitions (sometimes called genetic linkage in the Genetic Algorithm (GA) literature[8]) of the optimization problem. Messy genetic algorithms [4, 6, 5, 7, 10, 3], Dependency trees =-=[2]-=-, distribution estimation [15, 16], are some examples. For problems that can be decomposed into smaller subproblems in the chosen representation, these algorithms are likely to work effectively by red... |

97 | The gene expression messy genetic algorithm
- Kargupta
- 1996
(Show Context)
Citation Context ...ce. A detailed discussion on the need for addressing this class of problems on grounds of polynomial complexity search can be found elsewhere [10]. The Gene Expression Messy Genetic Algorithm (GEMGA) =-=[10, 11, 3]-=- offers one approach to detect decomposable partitions in terms of relations and classes. Regardless of whether or not the problem is decomposable, the GEMGA continues to look for relations that captu... |

53 | Locality-preserving hashing in multidimensional spaces
- Indyk, Motwani, et al.
- 1997
(Show Context)
Citation Context ...milarity among different chromosomes of the population. The current approach to do that requires O(m 2 ) pair-wise similarity computation, where m is the population size. Non-expansive hash functions =-=[14, 9]-=- can be used for doing this in O(1) time. In the recent past several mapping functions have been developed that map adjacent points in a domain (based on some distance metric) to adjacent points in th... |

31 | Binary and floating-point function optimization using messy genetic algorithms - Deb - 1991 |

31 | Non-expansive hashing
- Linial, Sasson
- 1996
(Show Context)
Citation Context ...milarity among different chromosomes of the population. The current approach to do that requires O(m 2 ) pair-wise similarity computation, where m is the population size. Non-expansive hash functions =-=[14, 9]-=- can be used for doing this in O(1) time. In the recent past several mapping functions have been developed that map adjacent points in a domain (based on some distance metric) to adjacent points in th... |

21 | Revisiting the GEMGA: Scalable evolutionary optimization through linkage learning
- Bandyopadhyay, Kargupta, et al.
- 1998
(Show Context)
Citation Context ...ary algorithms that inductively detect the decomposable partitions (sometimes called genetic linkage in the Genetic Algorithm (GA) literature[8]) of the optimization problem. Messy genetic algorithms =-=[4, 6, 5, 7, 10, 3]-=-, Dependency trees [2], distribution estimation [15, 16], are some examples. For problems that can be decomposed into smaller subproblems in the chosen representation, these algorithms are likely to w... |

18 |
distributions and graphical models in evolutionary optimization
- Schemata
- 1999
(Show Context)
Citation Context ...(sometimes called genetic linkage in the Genetic Algorithm (GA) literature[8]) of the optimization problem. Messy genetic algorithms [4, 6, 5, 7, 10, 3], Dependency trees [2], distribution estimation =-=[15, 16]-=-, are some examples. For problems that can be decomposed into smaller subproblems in the chosen representation, these algorithms are likely to work effectively by reducing the global search problem in... |

14 |
Learning Linkage to Efficiently Solve Problems of Bounded Difficulty Using Genetic Algorithms”, Illinois Genetic Algorithm
- Harik
- 1997
(Show Context)
Citation Context ...ary algorithms that inductively detect the decomposable partitions (sometimes called genetic linkage in the Genetic Algorithm (GA) literature[8]) of the optimization problem. Messy genetic algorithms =-=[4, 6, 5, 7, 10, 3]-=-, Dependency trees [2], distribution estimation [15, 16], are some examples. For problems that can be decomposed into smaller subproblems in the chosen representation, these algorithms are likely to w... |

9 |
Rapid, accurate optimizaiton of difficult problems using fast messy genetic algorithms
- Goldberg, Deb, et al.
- 1993
(Show Context)
Citation Context |

9 |
expression: The missing link in evolutionary computation
- Gene
- 2000
(Show Context)
Citation Context ...ge structure introduced by the optimization problem. At present, we are investigating this possibility using the theory of randomized near-isometric graph projection techniques. As reported elsewhere =-=[12]-=-, our early experiments on embedding an n-dimensional hamming space to an O(log 2 n) dimensional Euclidean space with a distortion bounded by O(log n) looks promising. As noted earlier, global detecti... |

6 |
Computational processes of evolution: The SEARCH perspective
- Kargupta
- 1996
(Show Context)
Citation Context |

1 | The genetic algorithms, linkage learning, and scalable data mining - Kargupta, Sanseverino, et al. - 1998 |