检索结果-内蒙古大学图书馆

evolutionary multiobjective optimization assisted by scalarization function approximation for high-dimensional expensive problems

引用

SWARM AND evolutionary COMPUTATION 2024年 86卷

作者： Horaguchi, Yuma Nishihara, Kei Nakata, Masaya Yokohama Natl Univ Fac Engn Tokiwadai 79-5 Yokohama Kanagawa 2408501 Japan

surrogate -assisted evolutionary algorithms (SAEAs) are a promising approach for solving expensive multiobjective optimization problems, but they often cannot address high -dimensional problems. Although one common approach to overcoming this challenge is to construct reliable surrogates, their accuracy inevitably deteriorates in a high -dimensional search space. Thus, this paper presents a novel SAEA based on scalarization function approximation, which is designed to strengthen its robustness against this deterioration. The proposed algorithm constructs an approximation model for each scalarization function defined in a decompositionbased framework. Each decomposed problem is then solved using multiple independent models trained for its neighbor problems. The intent is to decrease the risk of search performance degradations caused by unreliable approximations and retain the redundancy of the surrogate -assisted search to hedge the risk of over -fitting. Furthermore, each approximation model is adapted to a promising region of its corresponding decomposed problem to reduce the complexity of model fitting given a limited number of training samples. Experimental results show that the proposed algorithm is competitive with state-of-the-art SAEAs adapted for high -dimensional problems.

关键词： surrogate-assisted evolutionary algorithm Multiobjective optimization Scalarization function

来源：评论

学校读者我要写书评

暂无评论

Fitness Approximation Through Machine Learning with Dynamic Adaptation to the evolutionary State

引用

INFORMATION 2024年第12期15卷 744-744页

作者： Tzruia, Itai Halperin, Tomer Sipper, Moshe Elyasaf, Achiya Ben Gurion Univ Negev Dept Comp Sci IL-8410501 Beer Sheva Israel Ben Gurion Univ Negev Dept Software & Informat Syst Engn IL-8410501 Beer Sheva Israel

We present a novel approach to performing fitness approximation in genetic algorithms (GAs) using machine learning (ML) models, focusing on dynamic adaptation to the evolutionary state. We compare different methods for (1) switching between actual and approximate fitness, (2) sampling the population, and (3) weighting the samples. Experimental findings demonstrate significant improvement in evolutionary runtimes, with fitness scores that are either identical or slightly lower than those of the fully run GA-depending on the ratio of approximate-to-actual-fitness computation. Although we focus on evolutionary agents in Gymnasium (game) simulators-where fitness computation is costly-our approach is generic and can be easily applied to many different domains.

关键词： genetic algorithm machine learning fitness approximation surrogate-assisted evolutionary algorithm regression agent simulation

来源：评论

学校读者我要写书评

暂无评论

A Distributed RBF-assisted Differential Evolution for Distributed Expensive Constrained Optimization 1

引用

4th International Conference on Distributed Artificial Intelligence (DAI)

作者： Wei, Feng-Feng Guo, Xiao-Qi Qiu, Wen-Jin Chen, Tai-You Chen, Wei-Neng South China Univ Technol Guangzhou Peoples R China

ISBN: (数字)9783031255496

ISBN: (纸本)9783031255496;9783031255489

With the development of Internet of things and distributed computing techniques, distributed and expensive constrained optimization problems (DECOPs) have emerged in the industry. DECOPs refer to optimization problems with objective and constraint functions that are computationally expensive and can only be evaluated on multiple agents of distributed networks. In DECOPs, the raw data of each agent cannot be transmitted to other agents, but only objective or constraint value of a solution can be evaluated, resulting in the incomplete data on each agent. This paper proposes a distributed RBF-assisted differential evolution (DRADE) algorithm for solving DECOPs. In DRADE, we added a master agent to the distributed networks of DECOPs, connecting work agents that can evaluate objective or constraint values of candidate solutions to the master agent in a star topology. The proposed algorithm is composed of candidate generation and selection on master agent and radial basis function (RBF) management on work agents. In candidate generation and selection, differential evolution serves as an optimizer to generate candidate solutions assisted by RBF models received from work agents to replace expensive evaluations of candidate solutions in the master agent. In RBF management, each work agent constructs and updates a RBF model with its own data, which are updated by samples selected from candidate solutions received from the master agent and their expensively evaluated values. Statistical results and analysis of experiments carried out on benchmark test functions and engineering problems show that DRADE has superior performance than compared state-of-the-art SAEAs.

关键词： Distributed optimization surrogate-assisted evolutionary algorithm Expensive constrained optimization Differential evolution

来源：评论

学校读者我要写书评

暂无评论

surrogate-assisted Differential Evolution with Adaptation of Training Data Selection Criterion

Surrogate-assisted Differential Evolution with Adaptation of...

引用

IEEE Symposium Series on Computational Intelligence (IEEE SSCI)

作者： Nishihara, Kei Nakata, Masaya Yokohama Natl Univ Grad Sch Engn Sci Yokohama Kanagawa Japan

ISBN: (纸本)9781665487689

In surrogate-assisted evolutionary algorithms (SAEAs), the selection criterion of training data is a crucial option to improve the prediction accuracy of surrogate models. In this paper, we hypothesize that a proper selection criterion changes dependent on the problems and/or search situations. Accordingly, this paper proposes a surrogate-assisted differential evolution, which adapts the training data selection criterion to enhance the model accuracy and then optimization performance of our SAEA. In detail, the proposed algorithm builds multiple approximation models under different selection criteria. Then, the best-fitting model is utilized to screen candidate solutions. Experimental results show that the proposed algorithm outperforms the state-of-the-art SAEAs on a single-objective optimization benchmark suite up to 100 dimensions.

关键词： surrogate-assisted evolutionary algorithm adaptive data selection radial basis function differential evolution

来源：评论

学校读者我要写书评

暂无评论

A surrogate-assisted Partial Optimization for Expensive Constrained Optimization Problems 18th

A Surrogate-Assisted Partial Optimization for Expensive Cons...

引用

18th International Conference on Parallel Problem Solving from Nature (PPSN)

作者： Nishihara, Kei Nakata, Masaya Yokohama Natl Univ Yokohama Kanagawa 2408501 Japan

ISBN: (纸本)9783031700675;9783031700682

surrogate-assisted evolutionary algorithms (SAEAs) are gradually gaining attention as a method for solving expensive optimization problems with inequality constraints. Most SAEAs construct a surrogate model for each objective/constraint function and then aggregate approximation functions of constraints to estimate the feasibility of unevaluated solutions. However, because of the aggregation, the differences in the scales among constraints are ignored. Constraints with smaller scales do not benefit from constraint handling techniques as much as larger constraints, while the effects of handling constraints with larger scales scatter to the other many constraints. This results in an inefficient constraint optimization. Accordingly, this work proposes a new SAEA that partially optimizes each objective/constraint, namely surrogate-assisted partial optimization (SAPO). Solutions with better values of objective/constraint are selected from the evaluated solutions as the parent solutions and a focused objective/constraint is independently optimized using surrogate models one by one. Experimental results reveal the superiority of SAPO compared to the state-of-the-art SAEAs on a single-objective optimization problem suite with inequality constraints under an expensive optimization scenario.

关键词： surrogate-assisted evolutionary algorithm Constrained Optimization Problem Expensive Optimization Problem Radial Basis Function Network Differential Evolution

来源：评论

学校读者我要写书评

暂无评论

Progressive Sampling surrogate-assisted Particle Swarm Optimization for Large-Scale Expensive Optimization

Progressive Sampling Surrogate-Assisted Particle Swarm Optim...

引用

Genetic and evolutionary Computation Conference (GECCO)

作者： Wang, Hong-Rui Chen, Chun-Hua Li, Yun Zhang, Jun Zhi-Hui-Zhan South China Univ Technol Sch Comp Sci Engn Guangzhou 510006 Peoples R China South China Univ Technol Sch Software Engn Guangzhou 510006 Peoples R China Univ Elect & Technol China Shenzhen Inst Adv Study Shenzhen 518110 Peoples R China Hanyang Univ Anshan 15588 South Korea

ISBN: (纸本)9781450392372

surrogate-assisted evolutionary algorithms (SAEAs) have performed well on low- and medium-scale expensive optimization problems (EOPs). However, with the dimensionality increasing, existing SAEAs have trouble getting reliable surrogates for solving large-scale EOPs. In this paper, we propose a progressive sampling surrogate-assisted particle swarm optimization (PS-SAPSO) to efficiently solve the large-scale EOPs from the perspective of data collection and model training. For the data collection, a progressive sampling strategy with restart operation is proposed to collect the new sample solutions during the evolution process for training the radial basis function network (RBFN) surrogate. Specifically, a social learning particle swarm optimization is employed to generate the new sample solutions under the control of a progressively varying stop criterion. For the model training, a dynamic tunning strategy is proposed to obtain a reliable RBFN by adaptively adjusting the hyperparameter setting during the evolution process. The experimental result shows that PS-SAPSO can achieve competitive or better performance compared with four state-of-the-art SAEAs on widely used benchmark functions. Moreover, ablation experiments are conducted to show the effectiveness of the components of the PS-SAPSO algorithm.

关键词： Expensive optimization problem Large-scale problem surrogate-assisted evolutionary algorithm Particle swarm optimization Radial basis function network

来源：评论

学校读者我要写书评

暂无评论

A Constrained Sampling assisted Differential Evolution for Expensive Optimization 15

A Constrained Sampling Assisted Differential Evolution for E...

引用

15th International Conference on Advanced Computational Intelligence (ICACI)

作者： Wei, Feng-Feng Chen, Tai-You Shi, Xuan-Li Chen, Wei-Neng South China Univ Technol Sch Comp Sci & Engn Guangzhou Peoples R China

ISBN: (纸本)9798350321456

surrogate-assisted evolutionary algorithms (SAEAs) have become a widely used method to solve computationally expensive optimization problems, in which the evaluation of objective has high time cost. However, in many problems, the evaluation of constraints also has high time cost. These problems are known as expensive constrained optimization problems (ECOPs). To solve ECOPs, this paper proposes a constrained sampling assisted differential evolution (CSADE) algorithm. CSADE consists a global search process and a local search process. Global search process aims to explore the whole search space. During this process, constraints are transformed as a penalty to fitness and a radial basis function (RBF) model is built for the penalty fitness. DE/rand/1 algorithm is adopted for candidate generation to improve the population diversity. Local search process aims to exploit the promising areas. In this process, the objective and each constraint are approximated by an RBF model respectively. DE/best/1 algorithm is utilized for population convergence in promising areas. Extensive experiments with several SAEAs for ECOPs on benchmark test suites demonstrate the competitive performance of CSADE. Besides, comparison experiments on the optimization of bulk carrier design show the promising application in engineering optimization of CSADE.

关键词： differential evolution expensive constrained optimization surrogate-assisted evolutionary algorithm

来源：评论

学校读者我要写书评

暂无评论

Deadline-Driven Approach for Multi-Fidelity surrogate-assisted Environmental Model Calibration SWAN Wind Wave Model Case Study 19

Deadline-Driven Approach for Multi-Fidelity Surrogate-Assist...

引用

Genetic and evolutionary Computation Conference (GECCO)

作者： Nikitin, Nikolay O. Vychuzhanin, Pavel Hvatov, Alexander Deeva, Irina Kalyuzhnaya, Anna, V Kovalchuk, Sergey, V ITMO Univ St Petersburg Russia

ISBN: (纸本)9781450367486

This paper describes the approach for calibration of environmental models with the presence of time and quality restrictions. Advantages of the suggested strategy are based on two main concepts. The first advantage was provided by reducing the overall optimisation time due to the surrogate modelling of fitness function with the iterative gradual refinement of the environmental model fidelity (spatial and temporal resolution) for improving the fitness approximation. For the demonstration of the efficiency of surrogate-assisted multi-fidelity approach, it was compared with the baseline evolutionary calibration approach. The second advantage was assured by additional increasing of optimisation quality in the presence of strict deadline due to the building the strategy of multi-fidelity fitness approximation directly during the evolutionary algorithm execution. In order to prove the efficiency of the proposed dynamic strategy, it was compared with the preliminary meta-optimisation approach. As a case study, the wind wave model SWAN is used. The conducted experiments confirm the effectiveness of the proposed anytime approach and its applicability for the complex environmental models' parameters calibration.

关键词： deadline-driven optimisation surrogate-assisted evolutionary algorithm environmental model calibration SWAN model

来源：评论

学校读者我要写书评

暂无评论

MOEA/D-S³: MOEA/D using SVM-based surrogates adjusted to Subproblems for Many objective optimization

MOEA/D-S<SUP>3</SUP>: MOEA/D using SVM-based Surrogates adju...

引用

IEEE Congress on evolutionary Computation (CEC) as part of the IEEE World Congress on Computational Intelligence (IEEE WCCI)

作者： Sonoda, Takumi Nakata, Masaya Yokohama Natl Univ Collage Engn Sci Yokohama Kanagawa Japan

ISBN: (数字)9781728169293

ISBN: (纸本)9781728169293

This paper proposes a surrogate-assisted MOEA/D using SVM-based surrogates adjusted to subproblems (MOEA/D-S-3), which intends to achieve the following technical advantages. Firstly, in order to construct a proper surrogate while reducing learning cost to construct surrogates, a surrogate is an SVM-classifier that identifies a specific region of good solutions and thus its learning cost should be lower than a popular alternative approach, i.e., fitness approximation. Secondly, relying on the first advantage, multiple surrogates are constructed and each surrogate, like an expert, is adjusted to each subproblem defined in the MOEA/D framework in order to improve diversity and convergence of the Pareto set. Experimental results show that MOEA/D-S-3 outperforms MOEA/D on a number of many objective benchmark problems.

关键词： surrogate-assisted evolutionary algorithm many-objective optimization

来源：评论

学校读者我要写书评

暂无评论

Enhancing SAEAs with unevaluated solutions:a case study of relation model for expensive optimization

引用

Science China(Information Sciences) 2024年第2期67卷 41-58页

作者： Hao HAO Xiaoqun ZHANG Aimin ZHOU Institute of Natural Sciences Shanghai Jiao Tong University Shanghai Frontiers Science Center of Molecule Intelligent Syntheses School of Computer Science and Technology East China Normal University

surrogate-assisted evolutionary algorithms(SAEAs) hold significant importance in resolving expensive optimization problems. Extensive efforts have been devoted to improving the efficiency of SAEAs through the development of proficient model-assisted solution selection methods, which, however, first requires generating high-quality solutions. The fundamental paradigm of evaluating a limited number of solutions in each generation within SAEAs reduces the variance of adjacent populations, thereby affecting the quality of offspring solutions. This is a frequently encountered issue yet receives little attention. To address the issue, this paper presents a framework using unevaluated solutions to enhance the efficiency of *** surrogate model is employed to identify high-quality solutions for the direct generation of new solutions without evaluation. To ensure dependable selection, we have introduced two tailored relation models to select the optimal solution and the unevaluated population. A comprehensive experimental analysis is performed on two test suites, which showcases the superiority of the relation model over regression and classification models in the solution selection phase. Furthermore, the surrogate-selected unevaluated solutions with high potential have been shown to enhance the efficiency of the algorithm significantly.

关键词： expensive optimization unevaluated solutions relation model surrogate-assisted evolutionary algorithm

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：