Proceedings of the 11th Annual conference on Genetic and evolutionary computation最新文献

{"title":"Bloat control in genetic programming by evaluating contribution of nodes","authors":"A. Song, Dunhai Chen, Mengjie Zhang","doi":"10.1145/1569901.1570221","DOIUrl":"https://doi.org/10.1145/1569901.1570221","url":null,"abstract":"Unnecessary growth in program size is known as bloat problem in Genetic Programming. There are a large number of studies addressing this problem. In this paper, we propose an effective bloat control mechanism which is based on examining the contribution of each function node in the selected programs. Nodes without contribution will be removed before generating offspring. The results show that the method can significantly reduce program size without compromising fitness. Furthermore it speeds up evolution processes because of the saving in evaluation costs.","PeriodicalId":193093,"journal":{"name":"Proceedings of the 11th Annual conference on Genetic and evolutionary computation","volume":"129 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":"125711283","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":"Overlapped community detection in complex networks","authors":"C. Pizzuti","doi":"10.1145/1569901.1570019","DOIUrl":"https://doi.org/10.1145/1569901.1570019","url":null,"abstract":"Extracting and understanding community structure in complex networks is one of the most intensively investigated problems in recent years. In this paper we propose a genetic based approach to discover overlapping communities. The algorithm optimizes a fitness function able to identify densely connected groups of nodes by employing it on the line graph corresponding to the graph modelling the network. The method generates a division of the network in a number of groups in an unsupervised way. This number is automatically determined by the optimal value of the fitness function. Experiments on synthetic and real life networks show the capability of the method to successfully detect the network structure.","PeriodicalId":193093,"journal":{"name":"Proceedings of the 11th Annual conference on Genetic and evolutionary computation","volume":"17 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":"123779478","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":"Using crossover based similarity measure to improve genetic programming generalization ability","authors":"L. Vanneschi, Steven M. Gustafson","doi":"10.1145/1569901.1570054","DOIUrl":"https://doi.org/10.1145/1569901.1570054","url":null,"abstract":"Generalization is a very important issue in Machine Learning. In this paper, we present a new idea for improving Genetic Programming generalization ability. The idea is based on a dynamic two-layered selection algorithm and it is tested on a real-life drug discovery regression application. The algorithm begins using root mean squared error as fitness and the usual tournament selection. A list of individuals called ``repulsors'' is also kept in memory and initialized as empty. As an individual is found to overfit the training set, it is inserted into the list of repulsors. When the list of repulsors is not empty, selection becomes a two-layer algorithm: individuals participating to the tournament are not randomly chosen from the population but are themselves selected, using the average dissimilarity to the repulsors as a criterion to be maximized. Two kinds of similarity/dissimilarity measures are tested for this aim: the well known structural (or edit) distance and the recently defined subtree crossover based similarity measure. Although simple, this idea seems to improve Genetic Programming generalization ability and the presented experimental results show that Genetic Programming generalizes better when subtree crossover based similarity measure is used, at least for the test problems studied in this paper.","PeriodicalId":193093,"journal":{"name":"Proceedings of the 11th Annual conference on Genetic and evolutionary computation","volume":"38 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":"122845381","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":"Evaluation of population partitioning schemes in bayesian classifier EDAs: estimation of distribution algoithms","authors":"David Wallin, C. Ryan","doi":"10.1145/1569901.1569966","DOIUrl":"https://doi.org/10.1145/1569901.1569966","url":null,"abstract":"Several algorithms within the field of Evolutionary Computation have been proposed that effectively turn optimisation problems into supervised learning tasks. Typically such hybrid algorithms partition their populations into three subsets, high performing, low performing and mediocre, where the subset containing mediocre candidates is discarded from the phase of model construction. In this paper we will empirically compare this traditional partitioning scheme against two alternative schemes on a range of difficult problems from the literature. The experiments will show that at small population sizes, using the whole population is often a better approach than the traditional partitioning scheme, but partitioning around the midpoint and ignoring candidates at the extremes, is often even better.","PeriodicalId":193093,"journal":{"name":"Proceedings of the 11th Annual conference on Genetic and evolutionary computation","volume":"218 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":"131404614","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":"Initial-population bias in the univariate estimation of distribution algorithm","authors":"M. Pelikán, K. Sastry","doi":"10.1145/1569901.1569961","DOIUrl":"https://doi.org/10.1145/1569901.1569961","url":null,"abstract":"This paper analyzes the effects of an initial-population bias on the performance of the univariate marginal distribution algorithm (UMDA). The analysis considers two test problems: (1) onemax and (2) noisy onemax. Theoretical models are provided and verified with experiments. Intuitively, biasing the initial population toward the global optimum should improve performance of UMDA, whereas biasing the initial population away from the global optimum should have the opposite effect. Both theoretical and experimental results confirm this intuition. Effects of mutation on performance of UMDA with initial-population bias are also investigated.","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":"115142534","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":"Combining evolution strategy and gradient descent method for discriminative learning of bayesian classifiers","authors":"Xuefeng Chen, Xiabi Liu, Yunde Jia","doi":"10.1145/1569901.1569972","DOIUrl":"https://doi.org/10.1145/1569901.1569972","url":null,"abstract":"The optimization method is one of key issues in discriminative learning of pattern classifiers. This paper proposes a hybrid approach of the Covariance Matrix Adaptation Evolution Strategy (CMA-ES) and the gradient decent method for optimizing Bayesian classifiers under the SOFT target based Max-Min posterior Pseudo-probabilities (Soft-MMP) learning framework. In our hybrid optimization approach, the weighted mean of the parent population in the CMA-ES is adjusted by exploiting the gradient information of objective function, based on which the offspring is generated. As a result, the efficiency and the effectiveness of the CMA-ES are improved. We apply the Soft-MMP with the proposed hybrid optimization approach to handwritten digit recognition. The experiments on the CENPARMI database show that our handwritten digit classifier outperforms other state-of-the-art techniques. Furthermore, our hybrid optimization approach behaved better than not only the single gradient decent method but also the single CMA-ES in the experiments.","PeriodicalId":193093,"journal":{"name":"Proceedings of the 11th Annual conference on Genetic and evolutionary computation","volume":"110 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":"132783403","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":"Proceedings of the 11th Annual conference on Genetic and evolutionary computation","authors":"Franz Rothlauf","doi":"10.1145/1569901","DOIUrl":"https://doi.org/10.1145/1569901","url":null,"abstract":"These proceedings contain the papers presented at the 11th Annual Genetic and Evolutionary Computation Conference (GECCO-2009), held in Montreal, Canada, July 8-12, 2009. \u0000 \u0000After 2007, when GECCO was held in London, UK, this is the second time GECCO has been held outside the U.S. The generally high number of submissions of previous events has been maintained: 531 papers have been submitted for review, which is an increase of about 18% when compared to last year. Of these 531 papers, 220 were accepted as eight-page publications and 25 minutes presentations at the conference, yielding an acceptance ratio of 41,4%. In addition, 137 submissions (25,8%) have been accepted for poster presentations with two-page abstracts included in the proceedings. Last year, GECCO successfully moved over to electronic proceedings, and we continued with this publishing strategy as it greatly facilitates the handling of all conference materials. \u0000 \u0000GECCO has lived up to its motto of one conference, many mini-conferences. This year, there were 15 separate tracks that operated independently from each other. Each track had its own track chair(s) and individual program committee. A member of one track's program committee was not allowed to simultaneously be a member of another track's committee. To reduce any bias reviewers might have, all reviews were conducted double blind, no authors' names were included in the reviewed papers. About 600 researchers participated in the reviewing process. We want to thank them for all their work, which is highly appreciated and absolutely vital for the quality of the conference. \u0000 \u0000Track chairs have been asked to not accept more than 50% of their submissions as full papers. An appropriate acceptance rate is important in order to preserve the quality of the conference. Even though we were not bound by strong physical or environmental limitations on the number of accepted papers, we strove to keep our acceptance rate at the lower end. The scientific quality of the conference as well as that of the proceedings also is ensured by principles laid down in the GECCO by-laws of SIGEVO: (i) The GECCO conference shall be a broad-based conference encompassing the whole field of genetic and evolutionary computation. (ii) Papers will be published and presented as part of the main conference proceedings only after being peer reviewed. No invited papers shall be published (except for those of up to three invited plenary speakers). (iii) The peer review process shall be conducted consistent with the principle of division of powers performed by a multiplicity of independent program committees, each with expertise in the area of the paper being reviewed. (iv) The determination of the policy for the peer review process for each of the conference's independent program committees and the reviewing of papers for each program committee shall be performed by persons who occupy their positions by virtue of meeting objective and explicitly stated qualifications ","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":"133205899","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":"Learning complex robot control using evolutionary behavior based systems","authors":"Y. Kassahun, J. Schwendner, J. Gea, M. Edgington, F. Kirchner","doi":"10.1145/1569901.1569920","DOIUrl":"https://doi.org/10.1145/1569901.1569920","url":null,"abstract":"Evolving a monolithic solution for complex robotic problems is hard. One of the reasons for this is the difficulty of defining a global fitness function that leads to a solution with desired operating properties. The problem with a global fitness function is that it may not reward intermediate solutions that would ultimately lead to the desired operating properties. A possible way to solve such a problem is to decompose the solution space into smaller subsolutions with lower number of intrinsic dimensions. In this paper, we apply the design principles of behavior based systems to decompose a complex robot control task into subsolutions and show how to incrementally modify the fitness function that (1) results in desired operating properties as the subsolutions are learned, and (2) avoids the need to learn the coordination of behaviors separately. We demonstrate our method by learning to control a quadrocopter flying vehicle.","PeriodicalId":193093,"journal":{"name":"Proceedings of the 11th Annual conference on Genetic and evolutionary computation","volume":"2 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":"133300523","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":"SRaDE: an adaptive differential evolution based on stochastic ranking","authors":"Jinchao Liu, Zhun Fan, E. Goodman","doi":"10.1145/1569901.1570209","DOIUrl":"https://doi.org/10.1145/1569901.1570209","url":null,"abstract":"In this paper, we propose a methodology to improve the performance of the standard Differential Evolution (DE) in constraint optimization applications, in terms of accelerating its search speed, and improving the success rate. One critical mechanism embedded in the approach is applying Stochastic Ranking (SR) to rank the whole population of individuals with both objective value and constraint violation to be compared. The ranked population is then in a better shape to provide useful information e.g. direction to guide the search process. The strength of utilizing the directional information can be further controlled by a parameter - population partitioning factor, which is adjusted according to the evolution stage and generations. Because the adaptive adjustment of the parameter is predefined and does not need user input, the resulting algorithm is free of definition of this extra parameter and easier to implement. The performance of the proposed approach, which we call SRaDE (Stochastic Ranking based Adaptive Differential Evolution) is investigated and compared with standard DE. The experimental results show that SRDE significantly outperforms, or at least is comparable with standard DE in all the tested benchmark functions. We also conducted an experiment to compare SRaDE with SRDE - a variant of Stochastic Ranking based Differential Evolution without adaptive adjustment of the population partitioning factor. Experimental results show that SRaDE can also achieve improved performance over SRDE.","PeriodicalId":193093,"journal":{"name":"Proceedings of the 11th Annual conference on Genetic and evolutionary computation","volume":"15 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":"133630282","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 2: artificial life, evolutionary robotics, adaptive behavior, and evolvable hardware","authors":"R. Dreschler, Giovanni Squillero","doi":"10.1145/3257481","DOIUrl":"https://doi.org/10.1145/3257481","url":null,"abstract":"","PeriodicalId":193093,"journal":{"name":"Proceedings of the 11th Annual conference on Genetic and evolutionary computation","volume":"2 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":"131768471","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}