Article
Computer Science, Artificial Intelligence
Qiqi Liu, Yaochu Jin, Martin Heiderich, Tobias Rodemann
Summary: A new surrogate-assisted evolutionary algorithm is proposed in this study to handle expensive irregular multi-objective optimization problems. The algorithm balances convergence and diversity by adapting reference vectors and implementing a surrogate management strategy, effectively taking irregularity of the Pareto front into account.
KNOWLEDGE-BASED SYSTEMS
(2022)
Article
Computer Science, Artificial Intelligence
Jinglu Li, Peng Wang, Huachao Dong, Jiangtao Shen
Summary: This paper presents a two-stage surrogate-assisted evolutionary algorithm (TS-SAEA) for computationally expensive multi/many-objective optimization. The algorithm consists of a convergence stage and a diversity stage, which effectively optimize the objective space. Experimental results show that TS-SAEA has significant advantages on multi/many-objective optimization problems.
SWARM AND EVOLUTIONARY COMPUTATION
(2022)
Article
Computer Science, Information Systems
Yue Yang, Yongjie Ma, Minghao Wang, Peidi Wang
Summary: Evolutionary Algorithm is a mature global optimization method that effectively tackles complex problems. This paper proposes a dynamic multi-objective evolutionary algorithm based on gene sequencing and gene editing (GSGE) to optimize the Pareto optimal front (PF) distribution. The GSGE algorithm combines gene sequencing with support vector machine (SVM) design, stratified sampling, and multi-population strategy. Comparison experiments on 25 test functions demonstrate the significant advantages of GSGE.
INFORMATION SCIENCES
(2023)
Article
Computer Science, Artificial Intelligence
Fangfang Zhang, Yi Mei, Su Nguyen, Mengjie Zhang, Kay Chen Tan
Summary: This paper proposes a novel surrogate-assisted evolutionary multitask algorithm via GP to share useful knowledge between different scheduling tasks to improve training efficiency and effectiveness. Phenotypic characterization is used to measure the behaviors of scheduling rules and build a surrogate for each task. The proposed algorithm successfully improves the quality of scheduling heuristics for all scenarios.
IEEE TRANSACTIONS ON EVOLUTIONARY COMPUTATION
(2021)
Article
Computer Science, Information Systems
Raquel Espinosa, Fernando Jimenez, Jose Palma
Summary: Feature selection wrapper methods are powerful mechanisms for reducing the complexity of prediction models while preserving and even improving their precision. In this paper, a multi-surrogate assisted multi-objective evolutionary algorithm is proposed to improve generalization error. The experiments demonstrate the superiority of the proposed method over conventional wrapper methods using the same run times.
INFORMATION SCIENCES
(2023)
Article
Computer Science, Artificial Intelligence
Jinglu Li, Peng Wang, Huachao Dong, Jiangtao Shen, Caihua Chen
Summary: In this article, a classification surrogate-assisted multi-objective evolutionary algorithm (CSA-MOEA) is proposed for solving expensive optimization problems. The algorithm adopts a classification tree as the surrogate model and obtains valuable solutions through two infilling strategies.
KNOWLEDGE-BASED SYSTEMS
(2022)
Article
Computer Science, Artificial Intelligence
Yi Zhao, Jianchao Zeng, Ying Tan
Summary: The proposed method combines reference vector guided evolutionary algorithm and radial basis function networks to optimize individuals and introduces an infill strategy, showing competitive performance in solving computationally expensive many-objective optimization problems.
APPLIED SOFT COMPUTING
(2021)
Article
Energy & Fuels
Wenpeng Luan, Longfei Tian, Bochao Zhao
Summary: Dynamic tariffs are essential in demand response as they help smooth power consumption, reduce generation capacity requirement, and carbon emissions. However, existing works often overlook important factors such as user responses to tariffs when designing them. To address this issue, this paper proposes a new dynamic tariff design method that considers user responses to tariff changes. The method utilizes non-intrusive load monitoring technique to acquire information on rated power and user preferences for each appliance, which is then used to quantify user comfort or discomfort based on their appliance usage habits. A bi-level Stackelberg game model is then built to design optimal dynamic tariffs and simulate the impact of tariff changes on users' demand response plans. The results show that the proposed model generally outperforms benchmark methods in achieving peak shaving, low carbon emission, and user satisfaction.
Article
Computer Science, Artificial Intelligence
Xiwen Cai, Tao Zou, Liang Gao
Summary: This study proposes a surrogate-assisted multi-objective evolutionary algorithm that integrates multiple surrogate-assisted strategies to improve the optimization efficiency of computationally expensive multi-objective problems. The algorithm utilizes a surrogate-assisted penalty-based boundary intersection infill criterion and an operator-repeated offspring creation strategy for global search and diversity of Pareto optimal solutions. In addition, an improved surrogate-based multi-objective local search method is introduced to accelerate convergence speed. Experimental results demonstrate the superior performance of the proposed algorithm compared to state-of-the-art approaches.
APPLIED SOFT COMPUTING
(2023)
Article
Computer Science, Artificial Intelligence
Mengtian Wu, Lingling Wang, Jin Xu, Pengjie Hu, Pengcheng Xu
Summary: This paper proposes an adaptive technique for improving surrogate models and enhances the search performance of surrogate-assisted evolutionary algorithms by introducing multi-objective evolutionary algorithms as optimizers.
SWARM AND EVOLUTIONARY COMPUTATION
(2022)
Article
Computer Science, Information Systems
Hao Tong, Changwu Huang, Leandro L. Minku, Xin Yao
Summary: This paper provides a systematic review and comprehensive empirical study of surrogate models used in single-objective SAEAs, introducing a new taxonomy and comparing the characteristics of different models through experiments. The results are helpful for researchers to select suitable surrogate models when designing SAEAs.
INFORMATION SCIENCES
(2021)
Article
Computer Science, Artificial Intelligence
Mohammad A. Abido, Ashraf Elazouni
Summary: The study proposed a modified Multi-Objective Evolutionary Programming (MOEP) algorithm to solve scheduling problems of multi-mode activities, outperforming the benchmarked algorithms of SPEA-II and NSGA-II in terms of diversity and quality of the Pareto optimal set.
EXPERT SYSTEMS WITH APPLICATIONS
(2021)
Article
Computer Science, Artificial Intelligence
Ye Tian, Jiaxing Hu, Cheng He, Haiping Ma, Limiao Zhang, Xingyi Zhang
Summary: In this paper, a novel surrogate-assisted evolutionary algorithm is proposed, which employs a surrogate model to conduct pairwise comparisons between candidate solutions instead of directly predicting solutions' fitness values. Compared to regression and classification models, the proposed model based on pairwise comparison can better balance between positive and negative samples, and be directly used, reversely used, or ignored based on its reliability in model management. The experimental results on abundant benchmark and real-world problems demonstrate that the proposed surrogate model is more accurate and outperforms state-of-the-art surrogate models.
SWARM AND EVOLUTIONARY COMPUTATION
(2023)
Article
Computer Science, Artificial Intelligence
Qinghua Gu, Qian Wang, Neal N. Xiong, Song Jiang, Lu Chen
Summary: A surrogate-assisted evolutionary algorithm is proposed in this paper for solving expensive constrained multi-objective discrete optimization problems. By embedding random forest models and an improved stochastic ranking strategy, the algorithm makes significant progress in optimization efficiency and candidate solution quality.
COMPLEX & INTELLIGENT SYSTEMS
(2022)
Article
Computer Science, Information Systems
Maoqing Zhang, Lei Wang, Wuzhao Li, Bo Hu, Dongyang Li, Qidi Wu
Summary: This paper proposes a Many-Objective Evolutionary Algorithm with Adaptive Reference Vector (MaOEA-ARV) that can ensure both the spread and convergence of candidate solutions by dynamically adjusting reference vectors and adaptively partitioning candidate solutions into clusters. Experimental results demonstrate the effectiveness of MaOEA-ARV on test suites with up to 12 objectives.
INFORMATION SCIENCES
(2021)
Article
Computer Science, Artificial Intelligence
Meirong Chen, Yinan Guo, Yaochu Jin, Shengxiang Yang, Dunwei Gong, Zekuan Yu
Summary: This study proposes an environment-driven hybrid dynamic multi-objective evolutionary optimization method to balance the quality of obtained solutions and the computation cost, and select an appropriate optimization method based on the characteristics of the dynamic environment.
COMPLEX & INTELLIGENT SYSTEMS
(2023)
Article
Computer Science, Artificial Intelligence
Yan Zhou, Yaochu Jin, Yao Sun, Jinliang Ding
Summary: Liquid state machines (LSMs) are biologically more plausible than spiking neural networks for brain-inspired computing and neuromorphic engineering. However, optimizing and training complex recurrent network architectures in LSMs is challenging. This paper proposes a generative LSM with evolved reservoir architecture and optimized weights through a cooperative co-evolutionary algorithm and synaptic plasticity rules. Experimental results on benchmark problems show that the proposed algorithm outperforms state-of-the-art methods, and the data parallelism strategy effectively speeds up the evaluation process.
IEEE TRANSACTIONS ON EMERGING TOPICS IN COMPUTATIONAL INTELLIGENCE
(2023)
Article
Computer Science, Artificial Intelligence
Shangshang Yang, Haoyu Wei, Haiping Ma, Ye Tian, Xingyi Zhang, Yunbo Cao, Yaochu Jin
Summary: In this paper, a personalized exercise group assembly (PEGA) method is proposed to assemble personalized exercise groups based on students' abilities for flexible exercise recommendations. Experimental results demonstrate that the assembled exercises by the proposed method are more effective in enhancing students' proficiency on both poorly mastered and new knowledge concepts compared to existing exercise recommendation methods.
IEEE TRANSACTIONS ON EMERGING TOPICS IN COMPUTATIONAL INTELLIGENCE
(2023)
Editorial Material
Computer Science, Artificial Intelligence
Nian Zhang, Zhigang Zeng, Yaochu Jin
Summary: This special issue presents robust, explainable, and efficient next-generation deep learning algorithms with data privacy and theoretical guarantees to improve the understanding and explainability of deep neural networks; improve the accuracy of deep learning leveraging new stochastic optimization and neural architecture search; and increase the computational efficiency and stability of the deep learning training process with new algorithms that will scale.
KNOWLEDGE-BASED SYSTEMS
(2023)
Article
Computer Science, Artificial Intelligence
Xi Zhang, Guo Yu, Yaochu Jin, Feng Qian
Summary: This paper proposes a transfer learning based surrogate assisted evolutionary algorithm (TrSA-DMOEA) to efficiently solve expensive dynamic multi-objective optimization problems (EDMOPs). By utilizing the knowledge from previous high-quality solutions, Gaussian process models are built to improve the computational complexity and solution quality. Furthermore, an adaptive acquisition function based surrogate-assisted mechanism is introduced to balance convergence and diversity. Experimental results demonstrate the superiority of the proposed method in solving EDMOPs.
Article
Computer Science, Artificial Intelligence
Jia Liu, Ran Cheng, Yaochu Jin
Summary: This paper proposes a bi-fidelity multiobjective neural architecture search approach to enhance the adversarial robustness of deep neural networks. The approach formulates the neural architecture search problem as a multiobjective optimization problem and reduces the computational cost by combining three performance estimation methods. Extensive experiments on different datasets validate the effectiveness of the proposed approach.
Article
Computer Science, Artificial Intelligence
Xueming Yan, Yaochu Jin, Xiaohua Ke, Zhifeng Hao
Summary: Multi-echelon location-routing problems (ME-LRPs) are challenging due to uncertain assignment relationship and increasing number of echelons. This study utilizes a hierarchical fuzzy graph and evolutionary multi-tasking optimization algorithm to solve ME-LRPs. The proposed method considers multi-echelon assignment information and routing task selection to alleviate negative transfer between different tasks, and demonstrates competitiveness on benchmark problems.
COMPLEX & INTELLIGENT SYSTEMS
(2023)
Article
Computer Science, Artificial Intelligence
Xilu Wang, Yaochu Jin
Summary: In this paper, a personalized evolutionary Bayesian algorithm is proposed to handle optimization problems with personalized variables. This algorithm considers personalized information and measurement noise by using a contextual Gaussian process and an evolutionary algorithm.
COMPLEX & INTELLIGENT SYSTEMS
(2023)
Article
Computer Science, Artificial Intelligence
Yuanchao Liu, Jianchang Liu, Yaochu Jin, Fei Li, Tianzi Zheng
Summary: This work proposes a novel surrogate-assisted two-stage differential evolution (SA-TSDE) algorithm for expensive constrained optimization. It combines a hybrid differential evolution with a repair strategy in the first stage, and a clustering strategy with local surrogates in the second stage. Experimental results show that SA-TSDE is highly competitive compared with state-of-the-art methods.
IEEE TRANSACTIONS ON EMERGING TOPICS IN COMPUTATIONAL INTELLIGENCE
(2023)
Article
Computer Science, Artificial Intelligence
Rongsheng Wang, Qi Zhang, Xuewu Dai, Zhiming Yuan, Tao Zhang, Shuxin Ding, Yaochu Jin
Summary: This paper investigates the HSTR problem under a partial station blockage and proposes an efficient PS-SEGA algorithm for solving it. The algorithm utilizes permutation-based encoding and heuristic decoding methods to find an optimized rescheduled timetable. Additionally, a hybrid initialization method and restart strategy are presented to enhance the algorithm's performance.
APPLIED SOFT COMPUTING
(2023)
Article
Computer Science, Artificial Intelligence
Shuangming Yang, Yanwei Pang, Haowen Wang, Tao Lei, Jing Pan, Jian Wang, Yaochu Jin
Summary: In this article, a novel biologically plausible learning method is proposed to address the challenge of designing an efficient learning mechanism with spiking dendrites. The method utilizes a multi-scale learning rule with dendritic predictive characteristics and a two-phase learning mechanism based on burst-related plateau potential dynamics. Experimental results show that the proposed algorithm improves learning accuracy, reduces synaptic operations, and power consumption on neuromorphic hardware, while enhancing robustness and learning convergence speed.
Article
Computer Science, Artificial Intelligence
Jinbao Wang, Guoyang Xie, Yawen Huang, Jiayi Lyu, Feng Zheng, Yefeng Zheng, Yaochu Jin
Summary: Utilizing multi-modal neuroimaging data is effective in studying human cognitive activities and pathologies, but obtaining full sets of centrally collected paired data is impractical. Federated learning is needed to integrate dispersed data from different institutions. The proposed FedMed-GAN algorithm bridges the gap between federated learning and medical GAN, mitigating mode collapse without sacrificing generator performance. It outperforms state-of-the-art methods in comprehensive evaluations.
Article
Computer Science, Artificial Intelligence
Zhun Fan, Zhaojun Wang, Wenji Li, Xiaomin Zhu, Bingliang Hu, An-Min Zou, Weidong Bao, Minqiang Gu, Zhifeng Hao, Yaochu Jin
Summary: Swarm robotic systems (SRSs) are used in various fields, and designing local interaction rules for self-organization of robots is a challenging task. This study proposes a modular design automation framework for gene regulatory network (GRN) models that can generate entrapping patterns without the need for expertise. The framework utilizes basic network motifs and multi-objective genetic programming to optimize the structures and parameters of the GRN models. Simulation results show that the framework can generate novel GRN models with simpler structures and better performance in complex environments. Proof-of-concept experiments using e-puck robots validate the feasibility and effectiveness of the proposed GRN models.
SWARM AND EVOLUTIONARY COMPUTATION
(2023)
Article
Computer Science, Artificial Intelligence
Meng Wu, Xiaomin Zhu, Li Ma, Weidong Bao, Zhun Fan, Yaochu Jin
Summary: Multi-robot systems outperform single robots in accomplishing challenging tasks due to their properties that single robots lack. This paper proposes a cooperative hierarchical gene regulatory network (CH-GRN) to enhance mutual cooperation between robots and utilize obstacles for more effective target entrapment. The CH-GRN includes a target-neighbour-obstacle (TNO) pattern generation method and a concentration-vector method for adaptation and obstacle avoidance. Simulation experiments and Kilobot experiments demonstrate the effectiveness of the CH-GRN in various challenging environments with different types of obstacles.
SWARM AND EVOLUTIONARY COMPUTATION
(2023)
Article
Automation & Control Systems
Xilu Wang, Yaochu Jin, Sebastian Schmitt, Markus Olhofer
Summary: This study aims to address multi-objective optimization problems with multiple black-box and heterogeneous objectives. It proposes a multi-objective Bayesian evolutionary optimization (BEO) approach that alleviates search biases and achieves a balance between convergence and diversity. The proposed algorithm is able to find high-quality solutions for heterogeneous multi-objective optimization problems compared with state-of-the-art methods.
IEEE TRANSACTIONS ON SYSTEMS MAN CYBERNETICS-SYSTEMS
(2023)