Article
Computer Science, Information Systems
Guoyu Chen, Yinan Guo, Mingyi Huang, Dunwei Gong, Zekuan Yu
Summary: In this paper, a domain adaptation learning strategy based dynamic multiobjective evolutionary algorithm is proposed, which achieves efficient knowledge transfer by utilizing a mapping matrix and an increment information. Experimental results on 12 benchmark functions demonstrate its superior performance.
INFORMATION SCIENCES
(2022)
Article
Computer Science, Information Systems
Yulong Ye, Qiuzhen Lin, Lijia Ma, Ka-Chun Wong, Maoguo Gong, Carlos A. Coello Coello
Summary: This paper presents a multiple source transfer learning method, MSTL-DMOEA, for dynamic multiobjective optimization problems. By utilizing two transfer learning procedures, MSTL-DMOEA can fully exploit historical information and generate a higher-quality initial population. Experimental results demonstrate the superiority of MSTL-DMOEA in solving various types of problems.
INFORMATION SCIENCES
(2022)
Article
Computer Science, Artificial Intelligence
Li Yan, Wenlong Qi, A. K. Qin, Shengxiang Yang, Dunwei Gong, Boyang Qu, Jing Liang
Summary: In this paper, a manifold clustering-based predictor is proposed to reveal the underlying knowledge in the changing Pareto set (PS) and improve prediction accuracy. Experimental results show that the proposed algorithm outperforms state-of-the-art algorithms, especially on problems with nonlinear correlation between decision variables.
SWARM AND EVOLUTIONARY COMPUTATION
(2023)
Article
Automation & Control Systems
Xiao-Fang Liu, Jun Zhang, Jun Wang
Summary: This article presents a cooperative differential evolution algorithm with an attention-based prediction strategy for dynamic multiobjective optimization. Multiple populations are used to optimize multiple objectives and find subparts of the Pareto front. The algorithm achieves a balanced approximation of the Pareto front and adapts to changes in the environment by using a new attention-based prediction strategy. Experimental results demonstrate the superiority of the proposed method to state-of-the-art algorithms.
IEEE TRANSACTIONS ON SYSTEMS MAN CYBERNETICS-SYSTEMS
(2023)
Article
Computer Science, Artificial Intelligence
Ali Ahrari, Saber Elsayed, Ruhul Sarker, Daryl Essam, Carlos A. Coello Coello
Summary: Reinitialization approach involves prediction and variation operators for dynamic optimization. This study examines the impact of prediction accuracy and change severity on the optimal variation strength, and introduces an adaptive variation operator. Descriptive simulations were conducted to explore the method's learning capability and sensitivity to changes.
APPLIED SOFT COMPUTING
(2021)
Article
Automation & Control Systems
Xiaxia Li, Jingming Yang, Hao Sun, Ziyu Hu, Anran Cao
Summary: This study introduces a dual prediction strategy with inverse model (DPIM) to enhance the performance of multiobjective optimization problems in dynamic environments. Experimental results demonstrate that DPIM can achieve high-quality populations with good convergence and distribution in dynamic environments.
Article
Automation & Control Systems
Guoping Li, Yanmin Liu, Xicai Deng
Summary: Prediction-based methods have gained popularity for solving dynamic multiobjective optimization problems. However, most existing methods only utilize the optimal solutions from a few previous environments, neglecting earlier historical information and potentially reducing prediction accuracy. In this study, a novel prediction method based on fractional displacement is proposed to balance the accuracy and computation cost of prediction.
Article
Computer Science, Artificial Intelligence
Min Jiang, Zhenzhong Wang, Haokai Hong, Gary G. Yen
Summary: In the dynamic multiobjective optimization problems, utilizing the knee point transfer learning method KT-DMOEA can greatly improve computational efficiency and solution quality.
IEEE TRANSACTIONS ON EVOLUTIONARY COMPUTATION
(2021)
Article
Mathematics, Applied
Kejia He, Hongyu Cheng, Yuchen Zhou, Cuihua Xie
Summary: This study proposes a supply selection model that considers manufacturers' development and material ordering. It uses time series analysis to forecast the manufacturers' development trend and selects the appropriate protocol based on total material management cost. By using an evolutionary algorithm, a dynamic prediction protocol is obtained, allowing users to adapt to changing production goals in continuous time.
JOURNAL OF FUNCTION SPACES
(2022)
Article
Computer Science, Artificial Intelligence
Ying Chen, Juan Zou, Yuan Liu, Shengxiang Yang, Jinhua Zheng, Weixiong Huang
Summary: This paper proposes a new change response mechanism for dynamic multiobjective optimization problems (DMOPs) by combining a hybrid prediction strategy and a precision controllable mutation strategy. The hybrid prediction strategy enables quick adaptation to predictable environmental changes, while the precision controllable mutation strategy handles unpredictable environmental changes. The mechanism can adapt to various environmental changes and has been shown to be effective and competitive in experimental comparisons.
SWARM AND EVOLUTIONARY COMPUTATION
(2022)
Article
Computer Science, Artificial Intelligence
Wei Zhou, Liang Feng, Kay Chen Tan, Min Jiang, Yong Liu
Summary: Dynamic multiobjective optimization problem refers to a multiobjective optimization problem that varies over time. To solve this kind of problem, evolutionary search with prediction approaches have been developed to estimate the changes in the problem. However, existing prediction methods only focus on the change in the decision space. In this article, a new approach is proposed that conducts prediction from both the decision and objective spaces. Experimental results show the effectiveness of the proposed method in solving both benchmark and real-world DMOPs.
IEEE TRANSACTIONS ON EVOLUTIONARY COMPUTATION
(2022)
Article
Computer Science, Artificial Intelligence
Peidi Wang, Yongjie Ma
Summary: The DMOEA is a powerful solver for DMOPs, but the current algorithms lack strategies in both the environment response and static optimization stages. To address this, a new algorithm was proposed that incorporates different strategies in both stages to balance convergence and diversity. The algorithm uses nondominated solutions-guided evolution in the static optimization stage and fine prediction strategy in the environment response stage to improve performance in dynamic environments.
APPLIED INTELLIGENCE
(2023)
Article
Computer Science, Artificial Intelligence
Ziyu Hu, Zihan Li, Lixin Wei, Hao Sun, Xuemin Ma
Summary: Dynamic multiobjective optimization problems exist in daily life and industrial practice. The proposed algorithm, based on decision variable relationship, improves the performance by using different optimization methods for different types of decision variables.
NEURAL COMPUTING & APPLICATIONS
(2023)
Article
Computer Science, Information Systems
Yingjie Zou, Yuan Liu, Juan Zou, Shengxiang Yang, Jinhua Zheng
Summary: Sparse large scale multiobjective optimization problems (sparse LSMOPs) have a high degree of sparsity in the decision variables of their Pareto optimal solutions. Existing evolutionary algorithms for sparse LSMOPs fail to achieve sufficient sparsity due to inaccurate location of nonzero decision variables and lack of interaction between the locating process and optimizing process. To address this, a dynamic sparse grouping evolutionary algorithm (DSGEA) is proposed, which groups decision variables with comparable numbers of nonzero variables and applies improved evolutionary operators for optimization. DSGEA outperforms current EAs in experiments on real-world and benchmark problems, achieving sparser Pareto optimal solutions with precise locations of nonzero decision variables.
INFORMATION SCIENCES
(2023)
Article
Multidisciplinary Sciences
Suhaila Abd Halim, Yupiter H. P. Manurung, Muhamad Aiman Raziq, Cheng Yee Low, Muhammad Saufy Rohmad, John R. C. Dizon, Vladimir S. Kachinskyi
Summary: In this study, an application tool was developed using open-sourced and customized algorithm based on artificial neural networks to optimize resistance spot welding. The tool can predict the effects of welding parameters on tensile shear load bearing capacity and weld quality classifications with high accuracy. It provides a cost-effective and practical solution for small industries and research centers.
SCIENTIFIC REPORTS
(2023)
Article
Computer Science, Artificial Intelligence
Shufen Qin, Chaoli Sun, Qiqi Liu, Yaochu Jin
Summary: In surrogate-assisted multi-/many-objective evolutionary optimization, an efficient model management strategy is highly challenging due to the complex tradeoff between different objectives and accumulated uncertainty in the approximation of objective functions. This article proposes building surrogate models for each objective function and using a Gaussian process model to approximate a performance indicator. The experimental results demonstrate the competitiveness of the proposed method in comparison to state-of-the-art surrogate-assisted evolutionary algorithms.
IEEE TRANSACTIONS ON EVOLUTIONARY COMPUTATION
(2023)
Article
Computer Science, Artificial Intelligence
Ying Hu, Yong Zhang, Xiaozhi Gao, Dunwei Gong, Xianfang Song, Yinan Guo, Jun Wang
Summary: Feature selection is an important preprocessing technique in data mining and machine learning. This paper proposes a federated feature selection framework that introduces a trusted third participant to process and integrate optimal feature subsets from multiple participants. A federated evolutionary feature selection algorithm based on particle swarm optimization is proposed to effectively solve feature selection problems with multiple participants under privacy protection. Experimental results show that the proposed algorithm can significantly improve the classification accuracy of the feature subset selected by each participant while protecting data privacy.
KNOWLEDGE-BASED SYSTEMS
(2023)
Article
Automation & Control Systems
Qiqi Liu, Yaochu Jin, Martin Heiderich, Tobias Rodemann
Summary: Adapting the reference vectors in decomposition-based evolutionary many-objective optimization enhances solution diversity. However, it may slow down convergence. To address this issue, this study proposes a coordinated adaptation of reference vectors and scalarizing functions based on local angle thresholds and the age of the reference vectors. Experimental results show that this approach achieves a better balance between diversity and convergence for problems with regular and irregular Pareto fronts.
IEEE TRANSACTIONS ON SYSTEMS MAN CYBERNETICS-SYSTEMS
(2023)
Article
Automation & Control Systems
Ye Tian, Langchun Si, Xingyi Zhang, Kay Chen Tan, Yaochu Jin
Summary: This article proposes a novel PF estimation approach based on local models to address the difficulties in handling irregular PFs. By dividing the population into groups and building a local model for each group, the proposed approach can approximate PFs with complex geometrical structures. Experimental results show that the proposed algorithm outperforms the compared algorithms, especially on problems with highly irregular PFs.
IEEE TRANSACTIONS ON SYSTEMS MAN CYBERNETICS-SYSTEMS
(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)
Review
Computer Science, Artificial Intelligence
Chunlin He, Yong Zhang, Dunwei Gong, Xinfang Ji
Summary: This paper provides a systematic overview of surrogate-assisted evolutionary algorithms (SAEAs), including the necessity of studying SAEAs, commonly used surrogate models, classification and discussion of existing SAEAs, review of their applications in various fields, and suggestions for future research directions.
EXPERT SYSTEMS WITH APPLICATIONS
(2023)
Article
Computer Science, Artificial Intelligence
Xianfang Song, Yong Zhang, Dunwei Gong, Hui Liu, Wanqiu Zhang
Summary: This article proposes a hybrid feature selection algorithm using surrogate sample-assisted particle swarm optimization (SS-PSO), which divides the sample and feature spaces concurrently to reduce the computational cost and search space. Experimental results show that SS-PSO can obtain good feature subsets at the smallest computational cost on most datasets, making it a highly competitive method for high-dimensional feature selection.
IEEE TRANSACTIONS ON EVOLUTIONARY COMPUTATION
(2023)
Article
Computer Science, Artificial Intelligence
Yan Xiao, Yaochu Jin, Kuangrong Hao
Summary: This study focuses on few-shot relation classification (FSRC) and proposes an adaptive prototypical network and introduces a loss function for joint representation learning (JRL). The experiments show that this approach not only significantly improves accuracy but also enhances the generalization ability of FSRC.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2023)
Article
Computer Science, Artificial Intelligence
Cheng He, Lianghao Li, Ran Cheng, Yaochu Jin
Summary: This study proposes an efficient sampling-based method for large-scale multi-objective optimization, considering convergence enhancement, diversity maintenance, and local search. Experimental results demonstrate the effectiveness of the algorithm in solving large-scale multi-objective optimization problems.
COMPLEX & INTELLIGENT SYSTEMS
(2023)
Article
Computer Science, Artificial Intelligence
Shuai Wang, Beichen Ding, Yaochu Jin
Summary: This study develops a quantitative measure to tackle the influence maximization problem under structural failures in a complex network. It proposes a multi-factorial evolutionary algorithm, MFEARIM, to find seeds with robust influence ability. The competitive performance of MFEARIM is validated through experiments on synthetic and real-world networks, surpassing existing methods.
IEEE COMPUTATIONAL INTELLIGENCE MAGAZINE
(2023)
Article
Computer Science, Artificial Intelligence
Qiqi Liu, Felix Lanfermann, Tobias Rodemann, Markus Olhofer, Yaochu Jin
Summary: This paper formulates the building energy management problem as a 10-objective optimization problem and compares the performance of different algorithms. The experimental results show that the adaptive reference vector assisted algorithm is the most competitive, and the algorithms with surrogate assistance outperform those without surrogate in general.
IEEE COMPUTATIONAL INTELLIGENCE MAGAZINE
(2023)
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
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)
Article
Computer Science, Artificial Intelligence
Huiting Li, Yaochu Jin, Tianyou Chai
Summary: This article introduces a method for addressing the challenge in multi-objective Bayesian optimization of expensive problems by using multisource online transfer learning. The algorithm transfers knowledge from multiple computationally cheap problems by selecting sources and using style transfer mapping. It proposes an adaptive online multisource transfer learning method based on the relationship between the balance factor parameter and the transfer mapping method. Comparative studies confirm the effectiveness of the algorithm on various optimization benchmark problems and a real-world expensive optimization problem.
IEEE TRANSACTIONS ON EMERGING TOPICS IN COMPUTATIONAL INTELLIGENCE
(2023)
Article
Computer Science, Artificial Intelligence
Xueming Yan, Han Huang, Yaochu Jin, Liang Chen, Zhanning Liang, Zhifeng Hao
Summary: This article proposes a differential neural architecture search approach using multi-hashing embedding for multilingual text representation, which outperforms the state-of-the-art NAS methods for text classification.
IEEE TRANSACTIONS ON EMERGING TOPICS IN COMPUTATIONAL INTELLIGENCE
(2023)