There exit high variations among nano-devices in nano-electronic systems, owing to the extremely small size and the bottom-up self-assembly nanofabrication process. Therefore, it is important to develop logical function mapping techniques with the consideration of variation tolerance. In this paper, the variation tolerant logical mapping (VTLM) problem is treated as a multi-objective optimization problem (MOP), a hybridization of Non-dominated Sorting Genetic Algorithm II (NSGA-II) with a problem-specific local search is presented to solve the problem. The experiment results show that with the assistance of the problem-specific local search, the presented algorithm is effective, and can find better solutions than that without the local search.
{"title":"Hybridization of NSGA-II with greedy re-assignment for variation tolerant logic mapping on nano-scale crossbar architectures","authors":"Fugui Zhong, Bo Yuan, Bin Li","doi":"10.1145/2598394.2598430","DOIUrl":"https://doi.org/10.1145/2598394.2598430","url":null,"abstract":"There exit high variations among nano-devices in nano-electronic systems, owing to the extremely small size and the bottom-up self-assembly nanofabrication process. Therefore, it is important to develop logical function mapping techniques with the consideration of variation tolerance. In this paper, the variation tolerant logical mapping (VTLM) problem is treated as a multi-objective optimization problem (MOP), a hybridization of Non-dominated Sorting Genetic Algorithm II (NSGA-II) with a problem-specific local search is presented to solve the problem. The experiment results show that with the assistance of the problem-specific local search, the presented algorithm is effective, and can find better solutions than that without the local search.","PeriodicalId":298232,"journal":{"name":"Proceedings of the Companion Publication of the 2014 Annual Conference on Genetic and Evolutionary Computation","volume":"37 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2014-07-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128446635","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}
Learning Classifier Systems were introduced in the 1970s by John H. Holland as highly adaptive, cognitive systems. More than 40 years later, the introduction of Stewart W. Wilson's XCS, a highly engineered classifier system model, has transformed them into a state-of-the-art machine learning system. Learning classifier systems can effectively solve data-mining problems, reinforcement learning problems, and also cognitive, robotics control problems. In comparison to other, non-evolutionary machine learning techniques, their performance is competitive or superior, dependent on the setup and problem. Learning classifier systems can work both online and offline, they are extremely flexible, applicable to a larger range of problems, and are highly adaptive. Moreover, system knowledge can be easily extracted, visualized, or even used to focus the progressive search on particular interesting subspaces. This tutorial provides a gentle introduction to learning classifier systems and their general functionality. It then surveys the current theoretical understanding of the systems. Finally, we provide a suite of current successful LCS applications and discuss the most promising areas for future applications and research directions.
学习分类器系统是由John H. Holland在20世纪70年代引入的，是一种高度自适应的认知系统。40多年后，Stewart W. Wilson的XCS(一种高度工程化的分类器系统模型)的引入，将它们转变为最先进的机器学习系统。学习分类器系统可以有效地解决数据挖掘问题、强化学习问题以及认知、机器人控制问题。与其他非进化机器学习技术相比，它们的性能是有竞争力的还是更好的，这取决于设置和问题。学习分类器系统可以在线和离线工作，它们非常灵活，适用于更大范围的问题，并且具有高度的适应性。此外，系统知识可以很容易地提取、可视化，甚至用于将逐步搜索集中在特定感兴趣的子空间上。本教程提供了学习分类器系统及其一般功能的简单介绍。然后调查了当前对系统的理论认识。最后，我们提供了一套目前成功的LCS应用，并讨论了未来最有希望的应用领域和研究方向。
{"title":"Learning classifier systems: a gentle introduction","authors":"P. Lanzi","doi":"10.1145/2598394.2605343","DOIUrl":"https://doi.org/10.1145/2598394.2605343","url":null,"abstract":"Learning Classifier Systems were introduced in the 1970s by John H. Holland as highly adaptive, cognitive systems. More than 40 years later, the introduction of Stewart W. Wilson's XCS, a highly engineered classifier system model, has transformed them into a state-of-the-art machine learning system. Learning classifier systems can effectively solve data-mining problems, reinforcement learning problems, and also cognitive, robotics control problems. In comparison to other, non-evolutionary machine learning techniques, their performance is competitive or superior, dependent on the setup and problem. Learning classifier systems can work both online and offline, they are extremely flexible, applicable to a larger range of problems, and are highly adaptive. Moreover, system knowledge can be easily extracted, visualized, or even used to focus the progressive search on particular interesting subspaces. This tutorial provides a gentle introduction to learning classifier systems and their general functionality. It then surveys the current theoretical understanding of the systems. Finally, we provide a suite of current successful LCS applications and discuss the most promising areas for future applications and research directions.","PeriodicalId":298232,"journal":{"name":"Proceedings of the Companion Publication of the 2014 Annual Conference on Genetic and Evolutionary Computation","volume":"37 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2014-07-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133259028","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}
Optimizing several objectives that are often at odds with each other provides difficult challenges that are not encountered if having only one goal at hand. One intuitive way to solve a multi-objective problem is to aggregate the objectives and reformulate it as an optimization problem having just a single goal. This goal can be a designer specific aggregation of the objectives or a characterization of knees, trade-offs, utilities, stronger optimality concepts or preferences. This paper examines the theoretical relationships between two knee concepts and aggregate objective functions methods. The changes in the fitness landscape by utilizing different aggregations is also discussed.
{"title":"On the interrelationships between knees and aggregate objective functions","authors":"P. Shukla, M. Braun, H. Schmeck","doi":"10.1145/2598394.2598483","DOIUrl":"https://doi.org/10.1145/2598394.2598483","url":null,"abstract":"Optimizing several objectives that are often at odds with each other provides difficult challenges that are not encountered if having only one goal at hand. One intuitive way to solve a multi-objective problem is to aggregate the objectives and reformulate it as an optimization problem having just a single goal. This goal can be a designer specific aggregation of the objectives or a characterization of knees, trade-offs, utilities, stronger optimality concepts or preferences. This paper examines the theoretical relationships between two knee concepts and aggregate objective functions methods. The changes in the fitness landscape by utilizing different aggregations is also discussed.","PeriodicalId":298232,"journal":{"name":"Proceedings of the Companion Publication of the 2014 Annual Conference on Genetic and Evolutionary Computation","volume":"7 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2014-07-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133311718","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}
It is widely known that reinforcement learning is a more general problem than supervised learning. In fact, supervised learning can be seen as a class of reinforcement learning problems. However, only a couple of papers tested reinforcement learning algorithms in supervised learning problems. Here we propose a new and simpler way to abstract supervised learning for any reinforcement learning algorithm. Moreover, a new algorithm called Novelty-Organizing Classifiers is developed based on a Novelty Map population that focuses more on the novelty of the inputs than their frequency. A comparison of the proposed method with Self-Organizing Classifiers and BioHel on some datasets is presented. Even though BioHel is specialized in solving supervised learning problems, the results showed only a trade-off between the algorithms. Lastly, results on a maze problem validate the flexibility of the proposed algorithm beyond supervised learning problems. Thus, Novelty-Organizing Classifiers is capable of solving many supervised learning problems as well as a maze problem without changing any parameter at all. Considering the fact that no adaptation of parameters was executed, the proposed algorithm's basis seems interestingly flexible.
{"title":"Novelty-organizing classifiers applied to classification and reinforcement learning: towards flexible algorithms","authors":"Danilo Vasconcellos Vargas, H. Takano, J. Murata","doi":"10.1145/2598394.2598429","DOIUrl":"https://doi.org/10.1145/2598394.2598429","url":null,"abstract":"It is widely known that reinforcement learning is a more general problem than supervised learning. In fact, supervised learning can be seen as a class of reinforcement learning problems. However, only a couple of papers tested reinforcement learning algorithms in supervised learning problems. Here we propose a new and simpler way to abstract supervised learning for any reinforcement learning algorithm. Moreover, a new algorithm called Novelty-Organizing Classifiers is developed based on a Novelty Map population that focuses more on the novelty of the inputs than their frequency. A comparison of the proposed method with Self-Organizing Classifiers and BioHel on some datasets is presented. Even though BioHel is specialized in solving supervised learning problems, the results showed only a trade-off between the algorithms. Lastly, results on a maze problem validate the flexibility of the proposed algorithm beyond supervised learning problems. Thus, Novelty-Organizing Classifiers is capable of solving many supervised learning problems as well as a maze problem without changing any parameter at all. Considering the fact that no adaptation of parameters was executed, the proposed algorithm's basis seems interestingly flexible.","PeriodicalId":298232,"journal":{"name":"Proceedings of the Companion Publication of the 2014 Annual Conference on Genetic and Evolutionary Computation","volume":"31 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2014-07-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122374183","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}
M. Lones, Jane Elizabeth Alty, P. Duggan-Carter, A. J. Turner, D. R. S. Jamieson, S. Smith
Parkinson's disease is a chronic neurodegenerative condition that manifests clinically with various movement disorders. These are often treated with the dopamine-replacement drug levodopa. However, the dosage of levodopa must be kept as low as possible in order to avoid the drug's side effects, such as the involuntary, and often violent, muscle spasms called dyskinesia, or levodopa-induced dyskinesia. In this paper, we investigate the use of genetic programming for training classifiers that can monitor the effectiveness of levodopa therapy. In particular, we evolve classifiers that can recognise tremor and dyskinesia, movement states that are indicative of insufficient or excessive doses of levodopa, respectively. The evolved classifiers achieve clinically useful rates of discrimination, with AUC>0.9. We also find that temporal classifiers generally out-perform spectral classifiers. By using classifiers that respond to low-level features of the data, we identify the conserved patterns of movement that are used as a basis for classification, showing how this approach can be used to characterise as well as classify abnormal movement.
{"title":"Classification and characterisation of movement patterns during levodopa therapy for parkinson's disease","authors":"M. Lones, Jane Elizabeth Alty, P. Duggan-Carter, A. J. Turner, D. R. S. Jamieson, S. Smith","doi":"10.1145/2598394.2609852","DOIUrl":"https://doi.org/10.1145/2598394.2609852","url":null,"abstract":"Parkinson's disease is a chronic neurodegenerative condition that manifests clinically with various movement disorders. These are often treated with the dopamine-replacement drug levodopa. However, the dosage of levodopa must be kept as low as possible in order to avoid the drug's side effects, such as the involuntary, and often violent, muscle spasms called dyskinesia, or levodopa-induced dyskinesia. In this paper, we investigate the use of genetic programming for training classifiers that can monitor the effectiveness of levodopa therapy. In particular, we evolve classifiers that can recognise tremor and dyskinesia, movement states that are indicative of insufficient or excessive doses of levodopa, respectively. The evolved classifiers achieve clinically useful rates of discrimination, with AUC>0.9. We also find that temporal classifiers generally out-perform spectral classifiers. By using classifiers that respond to low-level features of the data, we identify the conserved patterns of movement that are used as a basis for classification, showing how this approach can be used to characterise as well as classify abnormal movement.","PeriodicalId":298232,"journal":{"name":"Proceedings of the Companion Publication of the 2014 Annual Conference on Genetic and Evolutionary Computation","volume":"388 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2014-07-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122859401","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}
Welcome to MedGEC 2014 MedGEC is the GECCO Workshop on the application of genetic and evolutionary computation (GEC) to problems in medicine and healthcare. A dedicated workshop at GECCO continues to provide a much needed focus for medical related applications of evolutionary computation, not only providing a clear definition of the state of the art, but also support to practitioners for whom genetic and evolutionary computation might not be their main area of expertise or experience. The Workshop has two main aims: To provide delegates with examples of the current state of the art of applications of GEC to medicine. To provide a forum in which researchers can discuss and exchange ideas, support and advise each other in theory and practice. This is the tenth year that this Workshop has been presented at GECCO, a reflection of the continued importance of genetic and evolutionary programming to medical applications. Presentations reflect awide range of healthcare areas including: Medical Imaging and Signal Processing; Data Mining Medical Data and Patient Records; Clinical Expert Systems and Knowledge-based Systems; Modeling and Simulation of Medical Processes; and Clinical Diagnosis and Therapy. Despite this broad range of applications, there are common themes of interest that are important in achieving a successful solution to the problems addressed. One recurring theme is accessibility to reliable medical datasets to train, test and evaluate genetic and evolutionary algorithms. It is acknowledged that achieving a "Gold Standard" dataset is problematic, due to the difficulties in gaining access to patients and the reliance on conventional clinical evaluation, which is often subjective, and therefore, potentially unreliable. A second important theme is the choice of genetic and evolutionary algorithm employed, its suitability for the problem at hand and the benefits of alternate representations. The Workshop provides a knowledgeable and supportive forum in which these and other issues can be discussed. Although traditionally a venue for practitioners in genetic and evolutionary computation development, it is hoped that the Workshop will also attract a wider audience, such as medical practitioners and other healthcare professionals who have an interest in the use of evolutionary algorithms in medicine. The Workshop organizers are always receptive to suggestions for new themes, areas for discussion and new activities and can be contacted directly via email.
{"title":"Session details: Workshop: medical applications of genetic and evolutionary computation","authors":"S. Smith, S. Cagnoni, R. Patton","doi":"10.1145/3250293","DOIUrl":"https://doi.org/10.1145/3250293","url":null,"abstract":"Welcome to MedGEC 2014 MedGEC is the GECCO Workshop on the application of genetic and evolutionary computation (GEC) to problems in medicine and healthcare. A dedicated workshop at GECCO continues to provide a much needed focus for medical related applications of evolutionary computation, not only providing a clear definition of the state of the art, but also support to practitioners for whom genetic and evolutionary computation might not be their main area of expertise or experience. The Workshop has two main aims: To provide delegates with examples of the current state of the art of applications of GEC to medicine. To provide a forum in which researchers can discuss and exchange ideas, support and advise each other in theory and practice. This is the tenth year that this Workshop has been presented at GECCO, a reflection of the continued importance of genetic and evolutionary programming to medical applications. Presentations reflect awide range of healthcare areas including: Medical Imaging and Signal Processing; Data Mining Medical Data and Patient Records; Clinical Expert Systems and Knowledge-based Systems; Modeling and Simulation of Medical Processes; and Clinical Diagnosis and Therapy. Despite this broad range of applications, there are common themes of interest that are important in achieving a successful solution to the problems addressed. One recurring theme is accessibility to reliable medical datasets to train, test and evaluate genetic and evolutionary algorithms. It is acknowledged that achieving a \"Gold Standard\" dataset is problematic, due to the difficulties in gaining access to patients and the reliance on conventional clinical evaluation, which is often subjective, and therefore, potentially unreliable. A second important theme is the choice of genetic and evolutionary algorithm employed, its suitability for the problem at hand and the benefits of alternate representations. The Workshop provides a knowledgeable and supportive forum in which these and other issues can be discussed. Although traditionally a venue for practitioners in genetic and evolutionary computation development, it is hoped that the Workshop will also attract a wider audience, such as medical practitioners and other healthcare professionals who have an interest in the use of evolutionary algorithms in medicine. The Workshop organizers are always receptive to suggestions for new themes, areas for discussion and new activities and can be contacted directly via email.","PeriodicalId":298232,"journal":{"name":"Proceedings of the Companion Publication of the 2014 Annual Conference on Genetic and Evolutionary Computation","volume":"6 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2014-07-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124204024","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}
E. Osaba, F. Díaz, R. Carballedo, Idoia De-la-Iglesia, E. Onieva, A. Perallos
This paper examines the influence of neutral crossover operators in a genetic algorithm (GA) applied to the one-dimensional bin packing problem. In the experimentation 16 benchmark instances have been used and the results obtained by three different GAs are compared with the ones obtained by an evolutionary algorithm (EA). The aim of this work is to determine whether an EA (with no crossover functions) can perform similarly to a GA.
{"title":"A study on the efficiency of neutral crossover operators in genetic algorithms applied to the bin packing problem","authors":"E. Osaba, F. Díaz, R. Carballedo, Idoia De-la-Iglesia, E. Onieva, A. Perallos","doi":"10.1145/2598394.2602268","DOIUrl":"https://doi.org/10.1145/2598394.2602268","url":null,"abstract":"This paper examines the influence of neutral crossover operators in a genetic algorithm (GA) applied to the one-dimensional bin packing problem. In the experimentation 16 benchmark instances have been used and the results obtained by three different GAs are compared with the ones obtained by an evolutionary algorithm (EA). The aim of this work is to determine whether an EA (with no crossover functions) can perform similarly to a GA.","PeriodicalId":298232,"journal":{"name":"Proceedings of the Companion Publication of the 2014 Annual Conference on Genetic and Evolutionary Computation","volume":"48 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2014-07-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124423212","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}
Taku Hasegawa, Kaname Matsumura, Kaiki Tsuchie, N. Mori, Keinosuke Matsumoto
Introducing the machine learning technique into evolutionary computation (EC) is one of the most important issues to expand EC design. In this paper, we proposed a novel method that combines the genetic algorithm and support vector machine to achieve the imaginary evolution without real fitness evaluations.
{"title":"Novel virtual fitness evaluation framework for fitness landscape learning evolutionary computation","authors":"Taku Hasegawa, Kaname Matsumura, Kaiki Tsuchie, N. Mori, Keinosuke Matsumoto","doi":"10.1145/2598394.2598496","DOIUrl":"https://doi.org/10.1145/2598394.2598496","url":null,"abstract":"Introducing the machine learning technique into evolutionary computation (EC) is one of the most important issues to expand EC design. In this paper, we proposed a novel method that combines the genetic algorithm and support vector machine to achieve the imaginary evolution without real fitness evaluations.","PeriodicalId":298232,"journal":{"name":"Proceedings of the Companion Publication of the 2014 Annual Conference on Genetic and Evolutionary Computation","volume":"36 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2014-07-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128914725","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}
Hadi Sharifi, Amin Nikanjam, Hossein Karshenas, Negar Najimi
Many of the real-world problems can be decomposed into a number of sub-problems for which the solutions can be found easier. However, proper decomposition of large problems remains a challenging issue, especially in optimization, where we need to find the optimal solutions more efficiently. Estimation of distribution algorithms (EDAs) are a class of evolutionary optimization algorithms that try to capture the interactions between problem variables when learning a probabilistic model from the population of candidate solutions. In this paper, we propose a type of synthesized problems, specially designed to challenge this specific ability of EDAs. They are based on the principal idea that each candidate solution to a problem may be simultaneously interpreted by two or more different structures where only one is true, resulting in the best solution to that problem. Of course, some of these structures may be more likely according to the statistics collected from the population of candidate solutions, but may not necessarily lead to the best solution. The experimental results show that the proposed benchmarks are indeed difficult for EDAs even when they use expressive models such as Bayesian networks to capture the interactions in the problem.
{"title":"Complexity of model learning in EDAs: multi-structure problems","authors":"Hadi Sharifi, Amin Nikanjam, Hossein Karshenas, Negar Najimi","doi":"10.1145/2598394.2598479","DOIUrl":"https://doi.org/10.1145/2598394.2598479","url":null,"abstract":"Many of the real-world problems can be decomposed into a number of sub-problems for which the solutions can be found easier. However, proper decomposition of large problems remains a challenging issue, especially in optimization, where we need to find the optimal solutions more efficiently. Estimation of distribution algorithms (EDAs) are a class of evolutionary optimization algorithms that try to capture the interactions between problem variables when learning a probabilistic model from the population of candidate solutions. In this paper, we propose a type of synthesized problems, specially designed to challenge this specific ability of EDAs. They are based on the principal idea that each candidate solution to a problem may be simultaneously interpreted by two or more different structures where only one is true, resulting in the best solution to that problem. Of course, some of these structures may be more likely according to the statistics collected from the population of candidate solutions, but may not necessarily lead to the best solution. The experimental results show that the proposed benchmarks are indeed difficult for EDAs even when they use expressive models such as Bayesian networks to capture the interactions in the problem.","PeriodicalId":298232,"journal":{"name":"Proceedings of the Companion Publication of the 2014 Annual Conference on Genetic and Evolutionary Computation","volume":"41 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2014-07-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129174221","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 tutorial gives a basic introduction to evolution strategies, a class of evolutionary algorithms. Key features such as mutation, recombination and selection operators are explained, and specifically the concept of self-adaptation of strategy parameters is introduced. All algorithmic concepts are explained to a level of detail such that an implementation of basic evolution strategies is possible. In addition, the tutorial also presents a brief taxonomy of contemporary evolution strategy variants, including e.g. the CMA-ES and variations thereof, and compares their performance for a small number of function evalutions - which represents many of today's practical application cases. Some guidelines for utilization as well as some application examples are also given.
{"title":"Introduction to evolution strategies","authors":"Thomas Bäck","doi":"10.1145/2598394.2605337","DOIUrl":"https://doi.org/10.1145/2598394.2605337","url":null,"abstract":"This tutorial gives a basic introduction to evolution strategies, a class of evolutionary algorithms. Key features such as mutation, recombination and selection operators are explained, and specifically the concept of self-adaptation of strategy parameters is introduced. All algorithmic concepts are explained to a level of detail such that an implementation of basic evolution strategies is possible. In addition, the tutorial also presents a brief taxonomy of contemporary evolution strategy variants, including e.g. the CMA-ES and variations thereof, and compares their performance for a small number of function evalutions - which represents many of today's practical application cases. Some guidelines for utilization as well as some application examples are also given.","PeriodicalId":298232,"journal":{"name":"Proceedings of the Companion Publication of the 2014 Annual Conference on Genetic and Evolutionary Computation","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2014-07-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116025435","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：