This paper describes a general scheme to convert sequential ant-based algorithms into parallel shared memory algorithms. The scheme is applied to an ant-based algorithm for the maximum clique problem. Extensive experimental results indicate that the parallel version provides noticeable improvements to the running time while maintaining comparable solution quality to that of the sequential version.
{"title":"Parallel shared memory strategies for ant-based optimization algorithms","authors":"T. N. Bui, ThanhVu Nguyen, Joseph R. Rizzo","doi":"10.1145/1569901.1569903","DOIUrl":"https://doi.org/10.1145/1569901.1569903","url":null,"abstract":"This paper describes a general scheme to convert sequential ant-based algorithms into parallel shared memory algorithms. The scheme is applied to an ant-based algorithm for the maximum clique problem. Extensive experimental results indicate that the parallel version provides noticeable improvements to the running time while maintaining comparable solution quality to that of the sequential version.","PeriodicalId":193093,"journal":{"name":"Proceedings of the 11th Annual conference on Genetic and evolutionary computation","volume":"24 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-07-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134599466","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Support Vector Machines (SVM) are very powerful classifiers in theory but their efficiency in practice rely on an optimal selection of hyper-parameters. A naïve or ad hoc choice of values for the latter can lead to poor performance in terms of generalization error and high complexity of parameterized models obtained in terms of the number of support vectors identified. This hyper-parameter estimation with respect to the aforementioned performance measures is often called the model selection problem in the SVM research community. In this paper we propose a strategy to select optimal SVM models in a dynamic fashion in order to attend that when knowledge about the environment is updated with new observations and previously parameterized models need to be re-evaluated, and in some cases discarded in favour of revised models. This strategy combines the power of the swarm intelligence theory with the conventional grid-search method in order to progressively identify and sort out potential solutions using dynamically updated training datasets. Experimental results demonstrate that the proposed method outperforms the traditional approaches tested against it while saving considerable computational time.
{"title":"A PSO-based framework for dynamic SVM model selection","authors":"Marcelo N. Kapp, R. Sabourin, P. Maupin","doi":"10.1145/1569901.1570066","DOIUrl":"https://doi.org/10.1145/1569901.1570066","url":null,"abstract":"Support Vector Machines (SVM) are very powerful classifiers in theory but their efficiency in practice rely on an optimal selection of hyper-parameters. A naïve or ad hoc choice of values for the latter can lead to poor performance in terms of generalization error and high complexity of parameterized models obtained in terms of the number of support vectors identified. This hyper-parameter estimation with respect to the aforementioned performance measures is often called the model selection problem in the SVM research community. In this paper we propose a strategy to select optimal SVM models in a dynamic fashion in order to attend that when knowledge about the environment is updated with new observations and previously parameterized models need to be re-evaluated, and in some cases discarded in favour of revised models. This strategy combines the power of the swarm intelligence theory with the conventional grid-search method in order to progressively identify and sort out potential solutions using dynamically updated training datasets. Experimental results demonstrate that the proposed method outperforms the traditional approaches tested against it while saving considerable computational time.","PeriodicalId":193093,"journal":{"name":"Proceedings of the 11th Annual conference on Genetic and evolutionary computation","volume":"7 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-07-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115554398","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Session details: Track 9: genetic algorithms","authors":"Jano von Hemert, T. Lenaerts","doi":"10.1145/3257488","DOIUrl":"https://doi.org/10.1145/3257488","url":null,"abstract":"","PeriodicalId":193093,"journal":{"name":"Proceedings of the 11th Annual conference on Genetic and evolutionary computation","volume":"21 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-07-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114769623","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Since the genomics revolution, bioinformatics has never been so popular. Many researchers have investigated with great success the use of evolutionary computation in bioinformatics [19] for example in the field of protein folding or determining genome sequences. In this paper, instead of using evolutionary computation as a way to provide new and innovative solutions to complex bioinformatics problems, we use genetic programming as a tool to evolve programs that are able to automatically classify research papers as dealing or not with a given protein. In a second part, we show that the attributes that are selected by the genetic programming evolved programs can be used efficiently for proteins classification.
{"title":"Genetic programming for protein related text classification","authors":"M. Segond, C. Fonlupt, D. Robilliard","doi":"10.1145/1569901.1570049","DOIUrl":"https://doi.org/10.1145/1569901.1570049","url":null,"abstract":"Since the genomics revolution, bioinformatics has never been so popular. Many researchers have investigated with great success the use of evolutionary computation in bioinformatics [19] for example in the field of protein folding or determining genome sequences. In this paper, instead of using evolutionary computation as a way to provide new and innovative solutions to complex bioinformatics problems, we use genetic programming as a tool to evolve programs that are able to automatically classify research papers as dealing or not with a given protein. In a second part, we show that the attributes that are selected by the genetic programming evolved programs can be used efficiently for proteins classification.","PeriodicalId":193093,"journal":{"name":"Proceedings of the 11th Annual conference on Genetic and evolutionary computation","volume":"46 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-07-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116712067","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Several practical problems in industry are difficult to optimize, both in terms of scalability and representation. Heuristics designed by domain experts are frequently applied to such problems. However, designing optimized heuristics can be a non-trivial task. One such difficult problem is the Facility Layout Problem (FLP) which is concerned with the allocation of activities to space. This paper is concerned with the block layout problem, where the activities require a fixed size and shape (modules). This problem is commonly divided into two sub problems; one of creating an initial feasible layout and one of improving the layout by interchanging the location of activities. We investigate how to extract novel heuristics for the FLP by applying an approach called Cooperative Coevolutionary Gene Expression Programming (CCGEP). By taking advantage of the natural problem decomposition, one species evolves heuristics for pre-scheduling, and another for allocating the activities onto the plant. An experimental, comparative approach investigates various features of the CCGEP approach. The results show that the evolved heuristics converge to suboptimal solutions as the problem size grows. However, coevolution has a positive effect on optimization of single problem instances. Expensive fitness evaluations may be limited by evolving generalized heuristics applicable to unseen fitness cases of arbitrary sizes.
{"title":"Scalability, generalization and coevolution -- experimental comparisons applied to automated facility layout planning","authors":"M. Furuholmen, K. Glette, M. Høvin, J. Tørresen","doi":"10.1145/1569901.1569997","DOIUrl":"https://doi.org/10.1145/1569901.1569997","url":null,"abstract":"Several practical problems in industry are difficult to optimize, both in terms of scalability and representation. Heuristics designed by domain experts are frequently applied to such problems. However, designing optimized heuristics can be a non-trivial task. One such difficult problem is the Facility Layout Problem (FLP) which is concerned with the allocation of activities to space. This paper is concerned with the block layout problem, where the activities require a fixed size and shape (modules). This problem is commonly divided into two sub problems; one of creating an initial feasible layout and one of improving the layout by interchanging the location of activities. We investigate how to extract novel heuristics for the FLP by applying an approach called Cooperative Coevolutionary Gene Expression Programming (CCGEP). By taking advantage of the natural problem decomposition, one species evolves heuristics for pre-scheduling, and another for allocating the activities onto the plant. An experimental, comparative approach investigates various features of the CCGEP approach. The results show that the evolved heuristics converge to suboptimal solutions as the problem size grows. However, coevolution has a positive effect on optimization of single problem instances. Expensive fitness evaluations may be limited by evolving generalized heuristics applicable to unseen fitness cases of arbitrary sizes.","PeriodicalId":193093,"journal":{"name":"Proceedings of the 11th Annual conference on Genetic and evolutionary computation","volume":"40 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-07-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115797390","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This paper presents a genetic programming-based symbolic regression approach to the construction of relational features in link analysis applications. Specifically, we consider the problems of predicting, classifying and annotating friends relations in friends networks, based upon features constructed from network structure and user profile data. We explain how the problem of classifying a user pair in a social network, as directly connected or not, poses the problem of selecting and constructing relevant features. We use genetic programming to construct features, represented by multiple symbol trees with base features as their leaves. In this manner, the genetic program selects and constructs features that may not have been originally considered, but possess better predictive properties than the base features. Finally, we present classification results and compare these results with those of the control and similar approaches.
{"title":"An evolutionary approach to constructive induction for link discovery","authors":"Tim Weninger, W. Hsu, Jing Xia, Waleed Aljandal","doi":"10.1145/1569901.1570248","DOIUrl":"https://doi.org/10.1145/1569901.1570248","url":null,"abstract":"This paper presents a genetic programming-based symbolic regression approach to the construction of relational features in link analysis applications. Specifically, we consider the problems of predicting, classifying and annotating friends relations in friends networks, based upon features constructed from network structure and user profile data. We explain how the problem of classifying a user pair in a social network, as directly connected or not, poses the problem of selecting and constructing relevant features. We use genetic programming to construct features, represented by multiple symbol trees with base features as their leaves. In this manner, the genetic program selects and constructs features that may not have been originally considered, but possess better predictive properties than the base features. Finally, we present classification results and compare these results with those of the control and similar approaches.","PeriodicalId":193093,"journal":{"name":"Proceedings of the 11th Annual conference on Genetic and evolutionary computation","volume":"31 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-07-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115490832","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Many objective optimization refers to optimization problems for which the number of objectives is significantly greater than conventionally studied 2 or 3. For such problems, large number of solutions become non-dominated, which reduces the convergence pressure of the Evolutionary Algorithms~(EAs) towards the Pareto Optimal Front. Recently, alternate secondary ranking schemes for have been suggested for NSGA-II in lieu of crowding distance to expedite its convergence for many objective problems. In this paper, we improvise upon an existing scheme~(epsilon dominance). The proposed approach is found to perform better than the other substitute distance assignment methods for the problems studied in this paper. A new diversity metric has also been proposed, which can be used in order to compare the performance of the various EAs.
{"title":"An improved secondary ranking for many objective optimization problems","authors":"H. Singh, A. Isaacs, T. Ray, W. Smith","doi":"10.1145/1569901.1570190","DOIUrl":"https://doi.org/10.1145/1569901.1570190","url":null,"abstract":"Many objective optimization refers to optimization problems for which the number of objectives is significantly greater than conventionally studied 2 or 3. For such problems, large number of solutions become non-dominated, which reduces the convergence pressure of the Evolutionary Algorithms~(EAs) towards the Pareto Optimal Front. Recently, alternate secondary ranking schemes for have been suggested for NSGA-II in lieu of crowding distance to expedite its convergence for many objective problems. In this paper, we improvise upon an existing scheme~(epsilon dominance). The proposed approach is found to perform better than the other substitute distance assignment methods for the problems studied in this paper. A new diversity metric has also been proposed, which can be used in order to compare the performance of the various EAs.","PeriodicalId":193093,"journal":{"name":"Proceedings of the 11th Annual conference on Genetic and evolutionary computation","volume":"53 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-07-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123430327","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Natural complex adaptive systems show many examples of self-organization and decentralization, such as pattern formation or swarm intelligence. Yet, only multicellular organisms possess the genuine architectural capabilities needed in many engineering application domains, from nanotechnologies to reconfigurable and swarm robotics. Biological development thus offers an important paradigm for a new breed of "evo-devo" computational systems. This work explores the evolutionary potential of an original multi-agent model of artificial embryogeny through differently parametrized simulations. It represents a rare attempt to integrate both self-organization and regulated architectures. Its aim is to illustrate how a developmental system, based on a truly indirect mapping from a modular genotype to a modular phenotype, can facilitate the generation of variations, thus structural innovation.
{"title":"Facilitating evolutionary innovation by developmental modularity and variability","authors":"R. Doursat","doi":"10.1145/1569901.1569996","DOIUrl":"https://doi.org/10.1145/1569901.1569996","url":null,"abstract":"Natural complex adaptive systems show many examples of self-organization and decentralization, such as pattern formation or swarm intelligence. Yet, only multicellular organisms possess the genuine architectural capabilities needed in many engineering application domains, from nanotechnologies to reconfigurable and swarm robotics. Biological development thus offers an important paradigm for a new breed of \"evo-devo\" computational systems. This work explores the evolutionary potential of an original multi-agent model of artificial embryogeny through differently parametrized simulations. It represents a rare attempt to integrate both self-organization and regulated architectures. Its aim is to illustrate how a developmental system, based on a truly indirect mapping from a modular genotype to a modular phenotype, can facilitate the generation of variations, thus structural innovation.","PeriodicalId":193093,"journal":{"name":"Proceedings of the 11th Annual conference on Genetic and evolutionary computation","volume":"41 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-07-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124483071","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
A key to the success of any genetic programming process is the use of a good alphabet of atomic building blocks from which solutions can be evolved efficiently. An alphabet that is too granular may generate an unnecessarily large search space; an inappropriately coarse grained alphabet may bias or prevent finding optimal solutions. Here we introduce a method that automatically identifies a small alphabet for a problem domain. We process solutions on the complexity-optimality Pareto front of a number of sample systems and identify terms that appear significantly more frequently than merited by their size. These terms are then used as basic building blocks to solve new problems in the same problem domain. We demonstrate this process on symbolic regression for a variety of physics problems. The method discovers key terms relating to concepts such as energy and momentum. A significant performance enhancement is demonstrated when these terms are then used as basic building blocks on new physics problems. We suggest that identifying a problem-specific alphabet is key to scaling evolutionary methods to higher complexity systems.
{"title":"Discovering a domain alphabet","authors":"Michael D. Schmidt, Hod Lipson","doi":"10.1145/1569901.1570047","DOIUrl":"https://doi.org/10.1145/1569901.1570047","url":null,"abstract":"A key to the success of any genetic programming process is the use of a good alphabet of atomic building blocks from which solutions can be evolved efficiently. An alphabet that is too granular may generate an unnecessarily large search space; an inappropriately coarse grained alphabet may bias or prevent finding optimal solutions. Here we introduce a method that automatically identifies a small alphabet for a problem domain. We process solutions on the complexity-optimality Pareto front of a number of sample systems and identify terms that appear significantly more frequently than merited by their size. These terms are then used as basic building blocks to solve new problems in the same problem domain. We demonstrate this process on symbolic regression for a variety of physics problems. The method discovers key terms relating to concepts such as energy and momentum. A significant performance enhancement is demonstrated when these terms are then used as basic building blocks on new physics problems. We suggest that identifying a problem-specific alphabet is key to scaling evolutionary methods to higher complexity systems.","PeriodicalId":193093,"journal":{"name":"Proceedings of the 11th Annual conference on Genetic and evolutionary computation","volume":"68 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-07-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124895390","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In this paper, we introduce Backward Time Related Association Rule Mining using Genetic Network Programming (GNP) with Database Rearrangement in order to find time related sequential association from time related databases effectively and efficiently. The proposed algorithm and experimental results are described using a traffic prediction problem.
{"title":"Backward time related association rule mining in trafficprediction using genetic network programming withdatabase rearrangement","authors":"Huiyu Zhou, S. Mabu, K. Shimada, K. Hirasawa","doi":"10.1145/1569901.1570223","DOIUrl":"https://doi.org/10.1145/1569901.1570223","url":null,"abstract":"In this paper, we introduce Backward Time Related Association Rule Mining using Genetic Network Programming (GNP) with Database Rearrangement in order to find time related sequential association from time related databases effectively and efficiently. The proposed algorithm and experimental results are described using a traffic prediction problem.","PeriodicalId":193093,"journal":{"name":"Proceedings of the 11th Annual conference on Genetic and evolutionary computation","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-07-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121611500","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：