Results 1  10
of
18
Using optimal dependencytrees for combinatorial optimization: Learning the structure of the search space
 Proceedings of the 14th International Conference on Machine Learning
, 1997
"... Many combinatorial optimization algorithms have no mechanism to capture interparameter dependencies. However, modeling such dependencies may allow an algorithm to concentrate its sampling more effectively on regions of the search space which have appeared promising in the past. We present an algori ..."
Abstract

Cited by 119 (2 self)
 Add to MetaCart
Many combinatorial optimization algorithms have no mechanism to capture interparameter dependencies. However, modeling such dependencies may allow an algorithm to concentrate its sampling more effectively on regions of the search space which have appeared promising in the past. We present an algorithm which incrementally learns secondorder probability distributions from good solutions seen so far, uses these statistics to generate optimal (in terms of maximum likelihood) dependency trees to model these distributions, and then stochastically generates new candidate solutions from these trees. We test this algorithm on a variety of optimization problems. Our results indicate superior performance over other tested algorithms that either (1) do not explicitly use these dependencies, or (2) use these dependencies to generate a more restricted class of dependency graphs. Scott Davies was supported by a Graduate Student Research Fellowship from the National Science Foundation. The views and conclusions contained in this document are those of the authors and should not be interpreted as representing the official policies, either expressed or implied of the National Science Foundation. Keywords:
Memetic Algorithms for Combinatorial Optimization Problems: Fitness Landscapes and Effective Search Strategies
, 2001
"... ..."
Realcoded Memetic Algorithms with crossover hillclimbing
 Evolutionary Computation
, 2004
"... This paper presents a realcoded memetic algorithm that applies a crossover hillclimbing to solutions produced by the genetic operators. On the one hand, the memetic algorithm provides global search (reliability) by means of the promotion of high levels of population diversity. On the other, the cro ..."
Abstract

Cited by 41 (9 self)
 Add to MetaCart
This paper presents a realcoded memetic algorithm that applies a crossover hillclimbing to solutions produced by the genetic operators. On the one hand, the memetic algorithm provides global search (reliability) by means of the promotion of high levels of population diversity. On the other, the crossover hillclimbing exploits the selfadaptive capacity of realparameter crossover operators with the aim of producing an effective local tuning on the solutions (accuracy). An important aspect of the memetic algorithm proposed is that it adaptively assigns different local search probabilities to individuals. It was observed that the algorithm adjusts the global/local search balance according to the particularities of each problem instance. Experimental results show that, for a wide range of problems, the method we propose here consistently outperforms other realcoded memetic algorithms which appeared in the literature.
Improving Genetic Algorithms by Search Space Reduction (with Applications to Flow Shop Scheduling
 GECCO99: Proceedings of the Genetic and Evolutionary Computation Conference
, 1999
"... Crossover operators that preserve common components can also preserve representation level constraints. Consequently, these constraints can be used to beneficially reduce the search space. For example, in flow shop scheduling problems with orderbased objectives (e.g. tardiness costs and earliness c ..."
Abstract

Cited by 9 (0 self)
 Add to MetaCart
(Show Context)
Crossover operators that preserve common components can also preserve representation level constraints. Consequently, these constraints can be used to beneficially reduce the search space. For example, in flow shop scheduling problems with orderbased objectives (e.g. tardiness costs and earliness costs), search space reductions have been implemented with precedence constraints. Experiments show that these (heuristically added) constraints can significantly improve the performance of Precedence Preserving Crossoveran operator which preserves common (orderbased) schemata. Conversely, the performance of Uniform OrderBased Crossover (the best traditional sequencing operator) improves lessit is based on combination. Overall, the results suggest that conditions exist where Precedence Preserving Crossover should be the best performing genetic sequencing operator. 1
Recombination without respect: Schema combination and disruption in genetic algorithm crossover
 Proceedings of the 2000 Genetic and Evolutionary Computation Conference
, 2000
"... Onepoint (or npoint) crossover has the property that schemata exhibited by both parents are ‘respected’⎯transferred to the offspring without disruption. In addition, new schemata may, potentially, be created by combination of the genes on which the parents differ. Some argue that the preservation ..."
Abstract

Cited by 7 (0 self)
 Add to MetaCart
(Show Context)
Onepoint (or npoint) crossover has the property that schemata exhibited by both parents are ‘respected’⎯transferred to the offspring without disruption. In addition, new schemata may, potentially, be created by combination of the genes on which the parents differ. Some argue that the preservation of similarity is the important aspect of crossover, and that the combination of differences (key to the buildingblock hypothesis) is unlikely to be valuable. In this paper, we discuss the operation of recombination on a hierarchical buildingblock problem. Uniform crossover, which preserves similarity, fails on this problem. Whereas, onepoint crossover, that both preserves similarity and combines differences, succeeds. In fact, a somewhat perverse recombination operator, that combines differences but destroys schemata that are common to both parents, also succeeds. Thus, in this problem, combination of schemata from dissimilar parents is required, and preserving similarity is not required. The test problem represents an extreme case, but it serves to illustrate the different aspects of recombination that are available in regular operators such as onepoint crossover. 1
Using A Priori Knowledge To Create Probabilistic Models For Optimization
 INTERNATIONAL JOURNAL OF APPROXIMATE REASONING
, 2002
"... Recent studies have examined the effectiveness of using probabilistic models to guide the sample generation process for searching high dimensional spaces. Although ..."
Abstract

Cited by 6 (0 self)
 Add to MetaCart
Recent studies have examined the effectiveness of using probabilistic models to guide the sample generation process for searching high dimensional spaces. Although
Introducing a New Advantage of Crossover: CommonalityBased Selection.” In these proceedings
, 1999
"... The CommonalityBased Crossover Framework defines crossover as a twostep process: 1) preserve the maximal common schema of two parents, and 2) complete the solution with a construction heuristic. In these “heuristic ” operators, the first step is a form of selection. This commonalitybased form of ..."
Abstract

Cited by 6 (4 self)
 Add to MetaCart
(Show Context)
The CommonalityBased Crossover Framework defines crossover as a twostep process: 1) preserve the maximal common schema of two parents, and 2) complete the solution with a construction heuristic. In these “heuristic ” operators, the first step is a form of selection. This commonalitybased form of selection has been isolated in GENIE. Using random parent selection and a nonelitist generational replacement scheme, GENIE does not include fitnessbased selection. However, a theoretical analysis shows that “ideal ” construction heuristics in GENIE can potentially converge to optimal solutions. Experimentally, results show that the effectiveness of practical construction heuristics can be amplified by commonalitybased restarts. Overall, it is shown that the commonality hypothesis is validschemata common to aboveaverage solutions are indeed above average. Since common schemata can only be identified by multiparent operators, commonalitybased selection is a unique advantage that crossover can enjoy over mutation. 1
Replacement strategies to preserve useful diversity in steadystate genetic algorithms
, 2005
"... In this paper, we propose a replacement strategy for steadystate genetic algorithms that considers two features of the candidate chromosome to be included into the population: a measure of the contribution of diversity to the population and the fitness function. In particular, the proposal tries to ..."
Abstract

Cited by 5 (0 self)
 Add to MetaCart
In this paper, we propose a replacement strategy for steadystate genetic algorithms that considers two features of the candidate chromosome to be included into the population: a measure of the contribution of diversity to the population and the fitness function. In particular, the proposal tries to replace an individual in the population with worse values for these two features. In this way, the diversity of the population becomes increased and the quality of the solutions gets better, thus preserving high levels of useful diversity. Experimental results show the proposed replacement strategy achieved significant performance for problems with different difficulties, which regards to other replacement strategies presented in the literature.
NonStandard Crossover for a Standard Representation
 CommonalityBased Feature Subset Selection,” GECCO99: Proceedings of the Genetic and Evolutionary Computation Conference
, 1999
"... The CommonalityBased Crossover Framework has been presented as a general model for designing problem specific operators. Following this model, the Common Features/Random Sample Climbing operator has been developed for feature subset selectiona binary string optimization problem. Although this pro ..."
Abstract

Cited by 5 (0 self)
 Add to MetaCart
(Show Context)
The CommonalityBased Crossover Framework has been presented as a general model for designing problem specific operators. Following this model, the Common Features/Random Sample Climbing operator has been developed for feature subset selectiona binary string optimization problem. Although this problem should be an ideal application for genetic algorithms with standard crossover operators, experiments show that the new operator can find better feature subsets for classifier training. 1
2000a, “Combination and Recombination in Genetic Algorithms
, 2000
"... Abstract. Recombination is supposed to enable the component characteristics from two parents to be extracted and then reassembled in different combinations – hopefully producing an offspring that has the good characteristics of both parents. This can work only if it is possible to identify which par ..."
Abstract

Cited by 3 (1 self)
 Add to MetaCart
(Show Context)
Abstract. Recombination is supposed to enable the component characteristics from two parents to be extracted and then reassembled in different combinations – hopefully producing an offspring that has the good characteristics of both parents. This can work only if it is possible to identify which parts of each parent should be extracted. Crossover in the standard GA, for example, takes subsets of genes that are adjacent on the genome. Other variations of the GA propose more sophisticated methods for identifying good subsets of genes within an individual. Our approach is different; rather than devising methods to enable successful extraction of gene subsets from the parents, we utilize variable size individuals which represent subsets of genes from the outset. By allowing each individual to represent a buildingblock explicitly, the normal action of selection can identify good buildingblocks without also promoting garbage genes. Then putting together two individuals (creating an offspring that is twice the size), straight forwardly produces the sum of the parents characteristics. This process is more properly combination than recombination since these building blocks have not been previously combined with any other. This paper summarizes our research on this approach and describes improved methods that reduce the domain knowledge required for successful application. 1