Pub Date : 2022-07-18DOI: 10.1109/CEC55065.2022.9870309
Kenjiro Takahashi, Y. Fukuyama, Shuhei Kawaguchi, Takaomi Sato
This paper proposes an optimal production scheduling method using the production simulator and multi-population global-best modified brain storm optimization (MP-GMBSO). Currently, in industry sector, decarbonization and carbon neutrality are approached by technical innovations such as Industry 4.0. In particular, optimal production scheduling researches which are important in production environments have been conducted actively. However, there is a gap between the previous optimal production scheduling researches and production schedule generating methods of practical production environments. The proposed method can fill the gap and it can be applied to the practical production environments. Results of the proposed method are compared with those of the conventional MBSO [7] and GMBSO based methods. It is verified that the proposed MP-GMBSO based method can find higher quality production schedules. In addition, it is verified that there is a significant difference among the conventional MBSO and GMBSO based methods, and the proposed MP-GMBSO based method with 0.05 significant level by the Friedman test as a priori test and the Wilcoxon signed rank test with Bonferroni-Holm correction as a post hoc test. In addition, the objective function of the target production scheduling has needles and it is found that the problem is one of the challenging problems to be optimized. The proposed MP-GMBSO based method can solve the problem better than the conventional MBSO and GMBSO based methods even with the challenging characteristic of the problem.
{"title":"Optimal Production Scheduling using a Production Simulator and Multi-population Global-best Modified Brain Storm Optimization","authors":"Kenjiro Takahashi, Y. Fukuyama, Shuhei Kawaguchi, Takaomi Sato","doi":"10.1109/CEC55065.2022.9870309","DOIUrl":"https://doi.org/10.1109/CEC55065.2022.9870309","url":null,"abstract":"This paper proposes an optimal production scheduling method using the production simulator and multi-population global-best modified brain storm optimization (MP-GMBSO). Currently, in industry sector, decarbonization and carbon neutrality are approached by technical innovations such as Industry 4.0. In particular, optimal production scheduling researches which are important in production environments have been conducted actively. However, there is a gap between the previous optimal production scheduling researches and production schedule generating methods of practical production environments. The proposed method can fill the gap and it can be applied to the practical production environments. Results of the proposed method are compared with those of the conventional MBSO [7] and GMBSO based methods. It is verified that the proposed MP-GMBSO based method can find higher quality production schedules. In addition, it is verified that there is a significant difference among the conventional MBSO and GMBSO based methods, and the proposed MP-GMBSO based method with 0.05 significant level by the Friedman test as a priori test and the Wilcoxon signed rank test with Bonferroni-Holm correction as a post hoc test. In addition, the objective function of the target production scheduling has needles and it is found that the problem is one of the challenging problems to be optimized. The proposed MP-GMBSO based method can solve the problem better than the conventional MBSO and GMBSO based methods even with the challenging characteristic of the problem.","PeriodicalId":153241,"journal":{"name":"2022 IEEE Congress on Evolutionary Computation (CEC)","volume":"17 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115575706","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}
Pub Date : 2022-07-18DOI: 10.1109/CEC55065.2022.9870307
H. Ishibuchi, Yiming Peng, Lie Meng Pang
Multi-modal multi-objective optimization problems have multiple equivalent Pareto sets, each of which is mapped to the entire Pareto front. A number of multi-modal multi-objective algorithms have been proposed to find all equivalent Pareto sets. Their performance is evaluated by computational experiments on multi-modal multi-objective test problems. A common feature of those test problems is that a single point on the Pareto front in the objective space corresponds to multiple clearly separated Pareto optimal solutions in the decision space. In this paper, we propose a new type of multi-modal multi-objective test problems where a single point on the Pareto front corresponds to an infinite number of Pareto optimal solutions (i.e., a subset of the decision space). This means that the mapping from the Pareto set in the decision space to the Pareto front in the objective space is a set-to-point mapping. For example, all points on a line in the decision space are mapped to the same single point on the Pareto front. As a result, the dimensionality of the Pareto set is larger than that of the Pareto front. We examine the search behavior of multi-modal multi-objective algorithms using the proposed test problems. Some interesting observations are reported.
{"title":"Multi-Modal Multi-Objective Test Problems with an Infinite Number of Equivalent Pareto Sets","authors":"H. Ishibuchi, Yiming Peng, Lie Meng Pang","doi":"10.1109/CEC55065.2022.9870307","DOIUrl":"https://doi.org/10.1109/CEC55065.2022.9870307","url":null,"abstract":"Multi-modal multi-objective optimization problems have multiple equivalent Pareto sets, each of which is mapped to the entire Pareto front. A number of multi-modal multi-objective algorithms have been proposed to find all equivalent Pareto sets. Their performance is evaluated by computational experiments on multi-modal multi-objective test problems. A common feature of those test problems is that a single point on the Pareto front in the objective space corresponds to multiple clearly separated Pareto optimal solutions in the decision space. In this paper, we propose a new type of multi-modal multi-objective test problems where a single point on the Pareto front corresponds to an infinite number of Pareto optimal solutions (i.e., a subset of the decision space). This means that the mapping from the Pareto set in the decision space to the Pareto front in the objective space is a set-to-point mapping. For example, all points on a line in the decision space are mapped to the same single point on the Pareto front. As a result, the dimensionality of the Pareto set is larger than that of the Pareto front. We examine the search behavior of multi-modal multi-objective algorithms using the proposed test problems. Some interesting observations are reported.","PeriodicalId":153241,"journal":{"name":"2022 IEEE Congress on Evolutionary Computation (CEC)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115784323","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}
Pub Date : 2022-07-18DOI: 10.1109/CEC55065.2022.9870342
Tiago F.D. Pinheiro, S. V. Ravelo, L. Buriol
In this paper, we study the k-labeled spanning forest problem (kLSF). The input for this problem is an undirected graph with labeled edges and a positive integer k. The goal is to find a spanning forest of the graph with at most $k$ different labels associated with the edges, minimizing the number of components. kLSF finds practical applications in different scenarios related to networks design and telecommunications. Solving it may help to reduce the negative impact of electromagnetic fields exposure on the population health or to increase profits of internet management companies, among others. The interest in kLSF is not only practical but also theoretical since the problem generalizes the best-known NP-hard minimum labeling spanning tree problem (MLST). To approach kLSF, we propose a fix-and-optimize matheuristic that was tested over several instances, achieving high-quality solutions in reasonable computational time. When compared to the best-known algorithms in the literature, our matheuristic outperformed the other proposals in most cases, finding better solutions in less computational time for the most challenging instances.
{"title":"A fix-and-optimize matheuristic for the k-labelled spanning forest problem","authors":"Tiago F.D. Pinheiro, S. V. Ravelo, L. Buriol","doi":"10.1109/CEC55065.2022.9870342","DOIUrl":"https://doi.org/10.1109/CEC55065.2022.9870342","url":null,"abstract":"In this paper, we study the k-labeled spanning forest problem (kLSF). The input for this problem is an undirected graph with labeled edges and a positive integer k. The goal is to find a spanning forest of the graph with at most $k$ different labels associated with the edges, minimizing the number of components. kLSF finds practical applications in different scenarios related to networks design and telecommunications. Solving it may help to reduce the negative impact of electromagnetic fields exposure on the population health or to increase profits of internet management companies, among others. The interest in kLSF is not only practical but also theoretical since the problem generalizes the best-known NP-hard minimum labeling spanning tree problem (MLST). To approach kLSF, we propose a fix-and-optimize matheuristic that was tested over several instances, achieving high-quality solutions in reasonable computational time. When compared to the best-known algorithms in the literature, our matheuristic outperformed the other proposals in most cases, finding better solutions in less computational time for the most challenging instances.","PeriodicalId":153241,"journal":{"name":"2022 IEEE Congress on Evolutionary Computation (CEC)","volume":"85 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115791394","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}
Pub Date : 2022-07-18DOI: 10.1109/CEC55065.2022.9870225
Xiang Yu, Li Zhang, Mei Shen
Nantong blue calico is a kind of important intangible cultural heritages in China. To better safeguard and inherit it in a digital way, it is necessary to construct a large-scale dataset for Nantong blue calico. As so far, however, we could not find a public dataset for blue calico. The goal of this paper is to give a public image dataset which named $N$ tBC consisting of Nantong blue calico patterns and provide a baseline result for the recognition of Nantong blue calico patterns. In this paper, we perform several baseline experiments on the NtBC dataset, including handcrafted and deep feature based classification methods. we compare some handcrafted methods and four kinds of popular convolutional neural networks (CNNs), including ResNet-50, AlexNet, GoogLeNet-V1 and VGGNet-16. Experimental results show that ResNet-50 yields an accuracy of 93.8% in the recognition performance, which shows that it is efficient to classify blue calico patterns through deep learning methods. As a consequence, this result provides the current best baseline result for Nantong blue calico image recognition. We believe our $N$ tBC will facilitate future research on Chinese traditional patterns development, fine grained visual classification, and imbalanced learning fields. We make the dataset and pre-trained models publicly available at https://github.com/facebook/react.
{"title":"Nantong Blue Calico Image Dataset and Its Recognition","authors":"Xiang Yu, Li Zhang, Mei Shen","doi":"10.1109/CEC55065.2022.9870225","DOIUrl":"https://doi.org/10.1109/CEC55065.2022.9870225","url":null,"abstract":"Nantong blue calico is a kind of important intangible cultural heritages in China. To better safeguard and inherit it in a digital way, it is necessary to construct a large-scale dataset for Nantong blue calico. As so far, however, we could not find a public dataset for blue calico. The goal of this paper is to give a public image dataset which named $N$ tBC consisting of Nantong blue calico patterns and provide a baseline result for the recognition of Nantong blue calico patterns. In this paper, we perform several baseline experiments on the NtBC dataset, including handcrafted and deep feature based classification methods. we compare some handcrafted methods and four kinds of popular convolutional neural networks (CNNs), including ResNet-50, AlexNet, GoogLeNet-V1 and VGGNet-16. Experimental results show that ResNet-50 yields an accuracy of 93.8% in the recognition performance, which shows that it is efficient to classify blue calico patterns through deep learning methods. As a consequence, this result provides the current best baseline result for Nantong blue calico image recognition. We believe our $N$ tBC will facilitate future research on Chinese traditional patterns development, fine grained visual classification, and imbalanced learning fields. We make the dataset and pre-trained models publicly available at https://github.com/facebook/react.","PeriodicalId":153241,"journal":{"name":"2022 IEEE Congress on Evolutionary Computation (CEC)","volume":"453 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115934195","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}
Pub Date : 2022-07-18DOI: 10.1109/CEC55065.2022.9870318
Yi Wang, Tao Li, Xiaojie Liu, Jian Yao
Feature selection algorithms aim to improve the per-formance of machine learning algorithms by removing irrelevant and redundant features. Various feature selection algorithms have been proposed, but most of them select a global feature subset for characterizing the entire sample space. In contrast, this study proposes an efficient discrete clonal selection algorithm for local feature selection called DCSA-LFS with three features: (1) local sample behaviors are considered, and a local clustering-based evaluation criterion is used to select a distinct optimized feature subset for each different sample region; (2) an improved discrete clonal selection algorithm is proposed, which uses a differential evolution-based mutation operator to enhance the search capability of clonal selection algorithms; and (3) a two-part antibody representation is adopted to automatically adjust the weight-related parameter. Experimental results on twelve UCI datasets show that DCSA-LFS is competitive with traditional filter-based feature selection algorithms and a clonal selection algorithm-based local feature selection algorithm.
{"title":"A discrete clonal selection algorithm for filter-based local feature selection","authors":"Yi Wang, Tao Li, Xiaojie Liu, Jian Yao","doi":"10.1109/CEC55065.2022.9870318","DOIUrl":"https://doi.org/10.1109/CEC55065.2022.9870318","url":null,"abstract":"Feature selection algorithms aim to improve the per-formance of machine learning algorithms by removing irrelevant and redundant features. Various feature selection algorithms have been proposed, but most of them select a global feature subset for characterizing the entire sample space. In contrast, this study proposes an efficient discrete clonal selection algorithm for local feature selection called DCSA-LFS with three features: (1) local sample behaviors are considered, and a local clustering-based evaluation criterion is used to select a distinct optimized feature subset for each different sample region; (2) an improved discrete clonal selection algorithm is proposed, which uses a differential evolution-based mutation operator to enhance the search capability of clonal selection algorithms; and (3) a two-part antibody representation is adopted to automatically adjust the weight-related parameter. Experimental results on twelve UCI datasets show that DCSA-LFS is competitive with traditional filter-based feature selection algorithms and a clonal selection algorithm-based local feature selection algorithm.","PeriodicalId":153241,"journal":{"name":"2022 IEEE Congress on Evolutionary Computation (CEC)","volume":"56 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124348983","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}
Pub Date : 2022-07-18DOI: 10.1109/CEC55065.2022.9870229
Cyrill Baumann, A. Martinoli
The automatic design of well-performing robotic controllers is still an unsolved problem due to the inherently large parameter space and noisy, often hard-to-define performance metrics, especially when sequential tasks need to be accomplished. Distal control architectures, which combine pre-coded basic behaviors into a (probabilistic) finite state machine offer a promising solution to this problem. In this paper, we enhance a Mixed-Discrete Particle Swarm Optimization (MDPSO) algorithm with an Optimal Computing Budget Allocation (OCBA) scheme to automatically synthesize distal control architectures. We benchmark MDPSO-OCBA's performance against the original MDPSO as well as the Iterated F-Race (IRACE) and the Mesh Adaptive Direct Search (MADS) algorithms on both a benchmark function with different noise levels and design problems of distal control architectures. More specifically, we evaluate the algorithms using high-fidelity simulations in three increasingly challenging scenarios involving parallel and sequential tasks. Additionally, the best performing controller generated in simulation by each optimization algorithm is compared with a manually designed solution and validated with physical experiments. The analysis on the benchmark function with different noise levels demonstrates MDPSO-OCBA's high robustness to noise. The comparison on the robotic control design problems shows that, without any meta-parameter tuning, MDPSO-OCBA is able to generate the best performing control architectures overall, closely followed by IRACE. They significantly outperform MADS for the more complex and noisier scenarios, resulting in competitive controllers in comparison to the manually designed one.
{"title":"A Noise-Resistant Mixed-Discrete Particle Swarm Optimization Algorithm for the Automatic Design of Robotic Controllers","authors":"Cyrill Baumann, A. Martinoli","doi":"10.1109/CEC55065.2022.9870229","DOIUrl":"https://doi.org/10.1109/CEC55065.2022.9870229","url":null,"abstract":"The automatic design of well-performing robotic controllers is still an unsolved problem due to the inherently large parameter space and noisy, often hard-to-define performance metrics, especially when sequential tasks need to be accomplished. Distal control architectures, which combine pre-coded basic behaviors into a (probabilistic) finite state machine offer a promising solution to this problem. In this paper, we enhance a Mixed-Discrete Particle Swarm Optimization (MDPSO) algorithm with an Optimal Computing Budget Allocation (OCBA) scheme to automatically synthesize distal control architectures. We benchmark MDPSO-OCBA's performance against the original MDPSO as well as the Iterated F-Race (IRACE) and the Mesh Adaptive Direct Search (MADS) algorithms on both a benchmark function with different noise levels and design problems of distal control architectures. More specifically, we evaluate the algorithms using high-fidelity simulations in three increasingly challenging scenarios involving parallel and sequential tasks. Additionally, the best performing controller generated in simulation by each optimization algorithm is compared with a manually designed solution and validated with physical experiments. The analysis on the benchmark function with different noise levels demonstrates MDPSO-OCBA's high robustness to noise. The comparison on the robotic control design problems shows that, without any meta-parameter tuning, MDPSO-OCBA is able to generate the best performing control architectures overall, closely followed by IRACE. They significantly outperform MADS for the more complex and noisier scenarios, resulting in competitive controllers in comparison to the manually designed one.","PeriodicalId":153241,"journal":{"name":"2022 IEEE Congress on Evolutionary Computation (CEC)","volume":"31 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114675384","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}
Pub Date : 2022-07-18DOI: 10.1109/CEC55065.2022.9870412
Matheus Bernardelli de Moraes, G. P. Coelho
Many real-world optimization problems involve time-consuming fitness evaluation. To reduce the computational cost of expensive evaluations, researchers have been developing surrogate models to approximate the objective function values of unevaluated candidate solutions. However, most of the research has been developed for continuous optimization problems, while only a few of them address surrogate modeling for expensive multi-objective Combinatorial Optimization Problems (COPs). COPs have inherently different challenges than continuous optimization. For example, (i) many COPs have categorical and nominal decision variables; (ii) they often require the combination of both global and local search mechanisms; and (iii) some of them have constraints that make them NP-hard problems, which makes them even more difficult to solve with a reasonable number of fitness evaluations. To address these issues, this paper proposes a surrogate-assisted evolutionary algorithm that combines the decomposition-based algorithm MOEA/D, Tabu Local Search, and Random Forest as a surrogate model to approximate the objective function of unevaluated individuals on multi-objective COPs. Experiments were conducted on constrained and unconstrained well-known multi-objective combinatorial optimization benchmark problems. The experimental results demonstrate that the proposed design outperforms state-of-the-art algorithms without violating the restrictions in the number of objective function evaluations, which indicates that it may be suitable for real-world expensive multi-objective COPs.
{"title":"A Random Forest-Assisted Decomposition-Based Evolutionary Algorithm for Multi-Objective Combinatorial Optimization Problems","authors":"Matheus Bernardelli de Moraes, G. P. Coelho","doi":"10.1109/CEC55065.2022.9870412","DOIUrl":"https://doi.org/10.1109/CEC55065.2022.9870412","url":null,"abstract":"Many real-world optimization problems involve time-consuming fitness evaluation. To reduce the computational cost of expensive evaluations, researchers have been developing surrogate models to approximate the objective function values of unevaluated candidate solutions. However, most of the research has been developed for continuous optimization problems, while only a few of them address surrogate modeling for expensive multi-objective Combinatorial Optimization Problems (COPs). COPs have inherently different challenges than continuous optimization. For example, (i) many COPs have categorical and nominal decision variables; (ii) they often require the combination of both global and local search mechanisms; and (iii) some of them have constraints that make them NP-hard problems, which makes them even more difficult to solve with a reasonable number of fitness evaluations. To address these issues, this paper proposes a surrogate-assisted evolutionary algorithm that combines the decomposition-based algorithm MOEA/D, Tabu Local Search, and Random Forest as a surrogate model to approximate the objective function of unevaluated individuals on multi-objective COPs. Experiments were conducted on constrained and unconstrained well-known multi-objective combinatorial optimization benchmark problems. The experimental results demonstrate that the proposed design outperforms state-of-the-art algorithms without violating the restrictions in the number of objective function evaluations, which indicates that it may be suitable for real-world expensive multi-objective COPs.","PeriodicalId":153241,"journal":{"name":"2022 IEEE Congress on Evolutionary Computation (CEC)","volume":"40 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114895222","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}
Pub Date : 2022-07-18DOI: 10.1109/CEC55065.2022.9870372
R. Matousek, T. Hulka
This paper deals with an advanced adjustment of stabilization sequences for complex chaotic systems by means of meta-evolutionary approaches in the form of a preliminary study. In this study, a two-dimensional discrete-time dynamic system denoted as Duffing map, also called Holmes map, was used. In general, the Duffing oscillator model represents a real system in the field of nonlinear dynamics. For example, an excited model of a string choosing between two magnets. There are many articles on the stabilization of various chaotic maps, but attempts to stabilize the Duffing map, moreover, for higher orbits, are rather the exception. In the case of period four, this is a novelty. This paper presents several approaches to obtaining stabilizing perturbation sequences. The problem of stabilizing the Duffing map turns out to be difficult and is a good challenge for metaheuristic algorithms, and also as benchmark function. The first approach is the optimal parameterization of the ETDAS model using multi-restart Nelder-Mead (NM) algorithm na Genetic Algorithm (GA). The second approach is to use the symbolic regression procedure. A perturbation model is obtained using Genetic Programming (GP). The third approach is two-level optimization, where the best GP model is subsequently optimized using NM and GA algorithms. A novelty of the approach is also the effective use of the objective function, precisely in relation to the process of optimization of higher periodic paths.
{"title":"Stabilization of Higher Periodic Orbits of the Duffing Map using Meta-evolutionary Approaches: A Preliminary Study","authors":"R. Matousek, T. Hulka","doi":"10.1109/CEC55065.2022.9870372","DOIUrl":"https://doi.org/10.1109/CEC55065.2022.9870372","url":null,"abstract":"This paper deals with an advanced adjustment of stabilization sequences for complex chaotic systems by means of meta-evolutionary approaches in the form of a preliminary study. In this study, a two-dimensional discrete-time dynamic system denoted as Duffing map, also called Holmes map, was used. In general, the Duffing oscillator model represents a real system in the field of nonlinear dynamics. For example, an excited model of a string choosing between two magnets. There are many articles on the stabilization of various chaotic maps, but attempts to stabilize the Duffing map, moreover, for higher orbits, are rather the exception. In the case of period four, this is a novelty. This paper presents several approaches to obtaining stabilizing perturbation sequences. The problem of stabilizing the Duffing map turns out to be difficult and is a good challenge for metaheuristic algorithms, and also as benchmark function. The first approach is the optimal parameterization of the ETDAS model using multi-restart Nelder-Mead (NM) algorithm na Genetic Algorithm (GA). The second approach is to use the symbolic regression procedure. A perturbation model is obtained using Genetic Programming (GP). The third approach is two-level optimization, where the best GP model is subsequently optimized using NM and GA algorithms. A novelty of the approach is also the effective use of the objective function, precisely in relation to the process of optimization of higher periodic paths.","PeriodicalId":153241,"journal":{"name":"2022 IEEE Congress on Evolutionary Computation (CEC)","volume":"69 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124511744","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}
Pub Date : 2022-07-18DOI: 10.1109/CEC55065.2022.9870387
Nicolas Roy, Charlotte Beauthier, Alexandre Mayer
Heuristic optimization methods such as Particle Swarm Optimization (PSO) depend on their parameters to achieve good performance on a given class of problems. Some modifications of heuristic algorithms aim to adapt those parameters during the optimization process. We present a framework to design such adaptation strategies using continuous fuzzy feedback control. Our framework, which is not tied to a particular algorithm, provides us with a simple interface where probes are sampled in the optimization process and parameters are fed back. The process of turning probes into parameters uses fuzzy logic rule sets, where the design of rules aims to maximize performance on a training benchmark. This meta-optimization is achieved by a Bayesian Optimizer (BO) with a Gradient Boosted Regression Trees (GBRT) prior. The robustness of the control is also assessed on a validation benchmark.
{"title":"Setup of a New Adaptive Fuzzy Particle Swarm Optimization Algorithm","authors":"Nicolas Roy, Charlotte Beauthier, Alexandre Mayer","doi":"10.1109/CEC55065.2022.9870387","DOIUrl":"https://doi.org/10.1109/CEC55065.2022.9870387","url":null,"abstract":"Heuristic optimization methods such as Particle Swarm Optimization (PSO) depend on their parameters to achieve good performance on a given class of problems. Some modifications of heuristic algorithms aim to adapt those parameters during the optimization process. We present a framework to design such adaptation strategies using continuous fuzzy feedback control. Our framework, which is not tied to a particular algorithm, provides us with a simple interface where probes are sampled in the optimization process and parameters are fed back. The process of turning probes into parameters uses fuzzy logic rule sets, where the design of rules aims to maximize performance on a training benchmark. This meta-optimization is achieved by a Bayesian Optimizer (BO) with a Gradient Boosted Regression Trees (GBRT) prior. The robustness of the control is also assessed on a validation benchmark.","PeriodicalId":153241,"journal":{"name":"2022 IEEE Congress on Evolutionary Computation (CEC)","volume":"105 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124785506","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}
Pub Date : 2022-07-18DOI: 10.1109/CEC55065.2022.9870302
Ahmed Hassan, N. Pillay
The automated design of optimization techniques holds great promise for advancing state-of-the-art optimization techniques and it has already taken over the manual design by human experts in some problems. Genetic algorithms are one of the key approaches for tackling the automated design problem. Unfortunately, these algorithms may take several hours to run as the fitness evaluation involves solving some benchmark instances to determine the quality of a candidate configuration. In this paper, we hybridize a meta-genetic algorithm with reinforcement learning to automatically design genetic algorithms for the two-dimensional bin packing problem. The task of the meta-genetic algorithm is to search the configuration space of genetic algorithms and the task of reinforcement learning is to decide whether to evaluate a candidate configuration or not. Therefore, avoiding wasting the computational budget on poor configurations. The proposed hybrid and the meta-genetic algorithm without reinforcement learning produce solvers for the two-dimensional bin packing problem that are competitive with the state-of-the-art algorithms. However, the proposed hybrid consumes about 25% of the computational effort required by the meta-genetic algorithm without reinforcement learning.
{"title":"Hybridizing A Genetic Algorithm With Reinforcement Learning for Automated Design of Genetic Algorithms","authors":"Ahmed Hassan, N. Pillay","doi":"10.1109/CEC55065.2022.9870302","DOIUrl":"https://doi.org/10.1109/CEC55065.2022.9870302","url":null,"abstract":"The automated design of optimization techniques holds great promise for advancing state-of-the-art optimization techniques and it has already taken over the manual design by human experts in some problems. Genetic algorithms are one of the key approaches for tackling the automated design problem. Unfortunately, these algorithms may take several hours to run as the fitness evaluation involves solving some benchmark instances to determine the quality of a candidate configuration. In this paper, we hybridize a meta-genetic algorithm with reinforcement learning to automatically design genetic algorithms for the two-dimensional bin packing problem. The task of the meta-genetic algorithm is to search the configuration space of genetic algorithms and the task of reinforcement learning is to decide whether to evaluate a candidate configuration or not. Therefore, avoiding wasting the computational budget on poor configurations. The proposed hybrid and the meta-genetic algorithm without reinforcement learning produce solvers for the two-dimensional bin packing problem that are competitive with the state-of-the-art algorithms. However, the proposed hybrid consumes about 25% of the computational effort required by the meta-genetic algorithm without reinforcement learning.","PeriodicalId":153241,"journal":{"name":"2022 IEEE Congress on Evolutionary Computation (CEC)","volume":"8 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128487992","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：