Swarm and Evolutionary Computation最新文献

{"title":"Complexity, interpretability and robustness of GP-based feature engineering in remote sensing","authors":"João E. Batista ,&nbsp;Adam K. Pindur ,&nbsp;Ana I.R. Cabral ,&nbsp;Hitoshi Iba ,&nbsp;Sara Silva","doi":"10.1016/j.swevo.2024.101761","DOIUrl":"10.1016/j.swevo.2024.101761","url":null,"abstract":"<div><div>Feature engineering is a crucial step in machine learning that provides better data for the learning algorithms to induce robust models, and this effort should be adapted to the capabilities of each algorithm. For example, classifiers that do not perform data transformations (e.g., cluster-based) perform better when the different classes are separated, typically requiring preprocessed data. Other models (e.g., decision trees) can perform several splits in the feature space, easily obtaining perfect results in training data, but have a higher risk of overfitting with unprocessed data. We use the rbd-GP and M3GP genetic programming algorithms to induce new features based on the original features, to be used by shallow and deep decision tree and random forest models. M3GP is wrapped around a learning algorithm, using its performance as fitness. This way, the induced features are adapted to the classifier, allowing us to compare the complexity of the features induced for the different classifiers. We measure the complexity of the induced features using several structural and functional complexity metrics found in the literature, also proposing a new metric that measures the separability of classes in the feature space. Like other authors, we use complexity as an interpretability metric, selecting three models to discuss and validate based on their performance and size. We apply these methods to remote sensing classification problems and solve two tasks that are hard due to the high similarity between the land cover classes: detecting cocoa agroforest and forecasting forest degradation up to one year in the future.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"92 ","pages":"Article 101761"},"PeriodicalIF":8.2,"publicationDate":"2024-12-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142758829","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A Self-adaptive two stage iterative greedy algorithm based job scales for energy-efficient distributed permutation flowshop scheduling problem","authors":"Yang Yu ,&nbsp;Quan Zhong ,&nbsp;Liangliang Sun ,&nbsp;Yuyan Han ,&nbsp;Qichun Zhang ,&nbsp;Xuelei Jing ,&nbsp;Zhujun Wang","doi":"10.1016/j.swevo.2024.101777","DOIUrl":"10.1016/j.swevo.2024.101777","url":null,"abstract":"<div><div>The production form of distributed manufacturing combined with energy-efficient scheduling has attracted great attention. In this paper, the energy-efficient distributed permutation flowshop scheduling problem with sequence-dependent setup times (EEDPFSP/SDST) is studied with the criterion of minimizing the total flowtime (TF) and total energy consumption (TEC). Firstly, by changing the initial ordering rule according to the total flowtime, an improved multi-objective NEH heuristic is presented to generate better initial individuals. Secondly, by analyzing the feature of EEDPFSP/SDST, a self-adaptive two stage iterative greedy algorithm based on the job scales (SAIG_bjs) is proposed, which includes a self-adaptive local search according to the job scale, and the energy-saving strategy with the sequence-dependent setup times characteristics based on the time margin between adjacent operations on the same job. Finally, the extensive experiments are adopted to test the performance of the proposed algorithm, and the experimental results demonstrate that the proposed SAIG_bjs algorithm is superior to the other five well-known algorithms.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"92 ","pages":"Article 101777"},"PeriodicalIF":8.2,"publicationDate":"2024-11-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142745522","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Two-stage bidirectional coevolutionary algorithm for constrained multi-objective optimization","authors":"Shulin Zhao ,&nbsp;Xingxing Hao ,&nbsp;Li Chen ,&nbsp;Tingfeng Yu ,&nbsp;Xingyu Li ,&nbsp;Wei Liu","doi":"10.1016/j.swevo.2024.101784","DOIUrl":"10.1016/j.swevo.2024.101784","url":null,"abstract":"<div><div>Objective optimization and constraint satisfaction are two primary and conflicting tasks in solving constrained multi-objective optimization problems (CMOPs). To better trade off them, this paper proposes a two-stage bidirectional coevolutionary algorithm, termed C-TBCEA, for constrained multi-objective optimization. It consists of two stages, with each concentrating on specific targets, i.e., the first stage primarily focuses on objective optimization while the second stage focuses on constraint satisfaction by employing different evolutionary strategies at each stage. Via the synergy of the two stages, a dynamic trade-off between objective optimization and constraint satisfaction can be achieved, thus overcoming the distinctive challenges that may be encountered at different stages of evolution. In addition, to take advantage of both feasible and infeasible solutions, we employ two populations, i.e., the main population that stores the non-dominated feasible solutions and the archive population that maintains the informative infeasible solutions, to prompt the bidirectional coevolution of them. To validate the effectiveness of the proposed C-TBCEA, experiments are carried out on 6 CMOP test suites and 17 real-world CMOPs. The results demonstrate that the proposed algorithm is very competitive with 9 state-of-the-art constrained multi-objective optimization evolutionary algorithms (CMOEAs).</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"92 ","pages":"Article 101784"},"PeriodicalIF":8.2,"publicationDate":"2024-11-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142745521","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Bayesian network structure learning based on discrete artificial jellyfish search: Leveraging scoring and graphical properties","authors":"Xuchen Yan,&nbsp;Xiaoguang Gao,&nbsp;Zidong Wang,&nbsp;Qianglong Wang,&nbsp;Xiaohan Liu","doi":"10.1016/j.swevo.2024.101781","DOIUrl":"10.1016/j.swevo.2024.101781","url":null,"abstract":"<div><div>It is challenging to obtain a credible Bayesian network (BN) from data to represent uncertain knowledge. Current swarm intelligence optimization methods lack targeted improvements based on domain-specific knowledge, which limits the learning accuracy and convergence speed. To address this gap, we propose a novel discrete artificial jellyfish search method for structure learning that leverages the scoring and graphical properties of BNs. Inspired by scoring functions and equivalence classes, a directional crossover operator is designed to efficiently narrow the crossover range. Additionally, a bidirectional search operator uses score increment guidance during mutations. By incorporating adjacency matrix series, global loop finding and deleting operators are applied to identify and eliminate all the minimalist loops simultaneously. They can avoid omitting the optimal solution. The experimental results show that the proposed algorithm outperforms the existing state-of-the-art algorithms in the scoring and convergence speed, which achieves an effective integration of group intelligence and structure learning.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"92 ","pages":"Article 101781"},"PeriodicalIF":8.2,"publicationDate":"2024-11-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142720578","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A learning-based memetic algorithm for energy-efficient distributed flow-shop scheduling with preventive maintenance","authors":"Jingjing Wang,&nbsp;Honggui Han","doi":"10.1016/j.swevo.2024.101772","DOIUrl":"10.1016/j.swevo.2024.101772","url":null,"abstract":"<div><div>In manufacturing systems, implementing preventive maintenance (PM) is essential for ensuring sustainable production since the inevitable wear and tear of machines can significantly affect production efficiency. In today’s decentralized economy, distributed shop scheduling has emerged within the framework of distributed manufacturing to reduce costs, enhance efficiency, and strengthen competitiveness. Thus, this paper proposes a learning-based memetic algorithm (LMA) for addressing the energy-efficient distributed flow-shop scheduling problem with preventive maintenance (EDFSP-PM) to minimize both makespan and total energy consumption simultaneously. First, a mathematical model is formulated and encoding and decoding methods are developed to map solutions to schedules with consideration of PM operations. Second, two heuristics are employed to generate high-quality solutions and various problem-specific operators are designed for different sub-problems and objectives. Third, a hierarchical learning mechanism is proposed via employing multi-layer Q-learning to select appropriate operators for solutions with diverse characteristics. Fourth, a feedback learning mechanism with solution pool is devised to reintegrate solutions from the pool into the search process to enhance search efficiency. Finally, numerical experiments are conducted to verify the effectiveness of the designed mechanisms. The comparative results demonstrate superior performance of the proposed LMA in terms of convergence and diversity.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"92 ","pages":"Article 101772"},"PeriodicalIF":8.2,"publicationDate":"2024-11-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142720580","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Provable space discretization based evolutionary search for scalable multi-objective security games","authors":"Hong Qian ,&nbsp;Yu-Peng Wu ,&nbsp;Rong-Jun Qin ,&nbsp;Xin An ,&nbsp;Yi Chen ,&nbsp;Aimin Zhou","doi":"10.1016/j.swevo.2024.101770","DOIUrl":"10.1016/j.swevo.2024.101770","url":null,"abstract":"<div><div>In the field of security, multi-objective security games (MOSGs) allow defenders to simultaneously protect targets from multiple heterogeneous attackers. MOSGs aim to simultaneously maximize all the heterogeneous payoffs, e.g., life, money, and crime rate, without merging heterogeneous attackers. In real-world scenarios, the number of targets and heterogeneous attackers may exceed the capability of most existing state-of-the-art (SOTA) methods, i.e., MOSGs are limited by the issue of scalability. In fact, there is still a lack of algorithms to improve scalability while ensuring accuracy. To this end, this paper proposes a general framework named Space Discretization based Evolutionary Search (SDES) based on many/multi-objective evolutionary algorithms (MOEAs) to scale up MOSGs to large-scale targets and heterogeneous attackers. SDES consists of four consecutive key components, i.e., discretization, optimization, evaluation, and refinement. Specifically, SDES first discretizes the originally high-dimensional continuous solution space to the low-dimensional discrete one by the maximal indifference property in game theory. This property helps EAs bypass the high-dimensional step function and simplify the solution of large-scale MOSGs. Then, MOEAs are used for optimization in the low-dimensional discrete solution space to obtain a well-spaced Pareto front. To evaluate solutions, SDES restores solutions back to the original space via greedily optimizing a novel divergence measurement. Finally, the refinement in SDES boosts the optimization performance with acceptable cost. Theoretically, we prove the optimization consistency and convergence of SDES. Experiment results show that SDES is the first linear-time MOSG algorithm for both large-scale attackers and targets. SDES can solve up to 20 attackers and 100 targets MOSG problems, while SOTA methods can only solve up to 8 attackers and 25 targets. An ablation study verifies the necessity of all components in SDES.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"92 ","pages":"Article 101770"},"PeriodicalIF":8.2,"publicationDate":"2024-11-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142720581","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A survey on vehicle–drone cooperative delivery operations optimization: Models, methods, and future research directions","authors":"Jing Zhou ,&nbsp;Jin Yi ,&nbsp;Zhenyu Yang ,&nbsp;Huayan Pu ,&nbsp;Xinyu Li ,&nbsp;Jun Luo ,&nbsp;Liang Gao","doi":"10.1016/j.swevo.2024.101780","DOIUrl":"10.1016/j.swevo.2024.101780","url":null,"abstract":"<div><div>With the rise of technology and market demand, unmanned devices, particularly drones, are increasingly used in logistics due to their speed and cost-efficiency. However, the persistence of limiting factors such as battery life and payload capacity has rendered vehicle-drone cooperative delivery an emerging and promising research area. This paper compiles relevant literature on the cooperative operation of vehicles and drones, summarizing key research directions, which encompass a novel classification framework, commonly utilized mathematical models, and solutions. Additionally, we collected algorithmic test cases employed in current research. Finally, the paper analyzes the current research status and the existing challenges and provides suggestions for future research directions.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"92 ","pages":"Article 101780"},"PeriodicalIF":8.2,"publicationDate":"2024-11-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142720439","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Multi-Objective Optimization for Distributed Generator and Shunt Capacitor Placement Considering Voltage-Dependent Nonlinear Load Models","authors":"Nil Kamal Yadav,&nbsp;Soumyabrata Das","doi":"10.1016/j.swevo.2024.101782","DOIUrl":"10.1016/j.swevo.2024.101782","url":null,"abstract":"<div><div>This paper aims to optimize the placement and sizing of distributed generator (DG) and shunt capacitor (SC) to minimize various single and multi-objective problems. Initially, three single objectives are addressed: minimizing active power loss, minimizing total operating cost, and minimizing total voltage deviation. A new mathematical expression for total operating cost, based on installation, maintenance, and operation costs of DG and SC, is developed. The Competitive Swarm Optimizer (CSO) algorithm is utilized to solve the optimization problem. The results obtained using CSO are compared with several other algorithms, including Cuckoo Search, Jaya, Teaching Learning Based Optimization, Particle Swarm Optimization, and Genetic Algorithm. The comparative results demonstrate that the CSO algorithm outperforms these methodologies. Subsequently, three multi-objective problems are formulated: simultaneous minimization of active power loss and total voltage deviation, simultaneous minimization of active power loss and total operating cost, simultaneous minimization of active power loss, total voltage deviation, and total operating cost. Multi-objective CSO is used to obtain a set of non-dominated optimal solutions, providing more realistic results. The R-method is then applied to select the best compromise solution from these non-dominated solutions. The developed model demonstrates its ability to find optimal placements of DG and SC, offering superior results compared to existing approaches. The proposed methodology is validated on six different non-linear loads such as constant power, constant current, constant impedance, industrial, commercial, and residential loads, highlighting its effectiveness and tested on the IEEE-34 bus radial distribution system.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"92 ","pages":"Article 101782"},"PeriodicalIF":8.2,"publicationDate":"2024-11-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142701008","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A weighted knowledge extraction strategy for dynamic multi-objective optimization","authors":"Yingbo Xie ,&nbsp;Junfei Qiao ,&nbsp;Ding Wang","doi":"10.1016/j.swevo.2024.101773","DOIUrl":"10.1016/j.swevo.2024.101773","url":null,"abstract":"<div><div>Multi-objective evolutionary algorithms suffer from performance degradation when solving dynamic multi-objective optimization problems (DMOPs) with a new conditional configuration from scratch, which motivates the research on knowledge extraction. However, most knowledge extraction strategies only focus on obtaining effective information from a single knowledge source, while ignoring the useful information from other knowledge sources with similar properties. Motivated by this, a weighted multi-source knowledge extraction strategy-based dynamic multiobjective evolutionary algorithm is proposed. First, a similarity criterion based on angle information is constructed to quantify similarity between different source domains and the target domain. Second, a knowledge extraction technique is developed to select a specific number of individuals from each source domain using a distance metric. Third, a generation strategy based on dynamic weighting mechanism is proposed, which generates a certain number of individuals and merges these individuals into the initial population within the new environment. Finally, the comprehensive experiments are conducted on public DMOP benchmarks and demonstrate the devised method significantly outperforms the state-of-the-art competing algorithms.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"92 ","pages":"Article 101773"},"PeriodicalIF":8.2,"publicationDate":"2024-11-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142701007","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A multi-strategy self-adaptive differential evolution algorithm for assembly hybrid flowshop lot-streaming scheduling with component sharing","authors":"Yiling Lu ,&nbsp;Qiuhua Tang ,&nbsp;Shujun Yu ,&nbsp;Lixin Cheng","doi":"10.1016/j.swevo.2024.101783","DOIUrl":"10.1016/j.swevo.2024.101783","url":null,"abstract":"<div><div>Lot-streaming helps to achieve a more balanced utilization of parallel machines and more timely assembly of components, while component sharing increases the flexibility and commonality of assembly operations. Thus, this work addresses an assembly hybrid flowshop lot-streaming scheduling problem with component sharing. A mixed-integer linear programming model is formulated to scrutinize the coupling relations among variables <em>i.e.</em> sub-lot splitting, machine allocation, processing sequencing, and assembly sequencing, and to minimize the maximum completion time and work-in-process inventory lexicographically. To solve the above problem efficiently, a multi-strategy self-adaptive differential evolution (MSDE) algorithm is developed. In MSDE, three problem-specific strategies that consider component integrity and specific requirements of production and assembly are integrated to enhance the initial population in terms of diversity and solution quality. A Q-learning-based selection mechanism is proposed to self-adaptively select an appropriate combination from mutation and crossover operators for achieving a balance between exploration and exploitation. An inventory reduction strategy is appended to largely reduce work-in-process components without extending completion time. Four conclusions are drawn from extensive experiments: (1) The ensemble of three population initialization strategies is superior to each individual one; (2) The Q-learning-based optimizer selection is more effective and robust than the single optimizer-based one; (3) The work-in-process inventory reduction strategy demonstrates remarkable effectiveness for most solutions; (4) MSDE outperforms the existing state-of-the-art algorithms in most cases.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"92 ","pages":"Article 101783"},"PeriodicalIF":8.2,"publicationDate":"2024-11-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142701006","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}