Pub Date : 2022-07-18DOI: 10.1109/CEC55065.2022.9870277
Muhammad Hassaan Farooq Butt, Hamail Ayaz, Muhammad Ahmad, J. Li, R. Kuleev
In forensic sciences, blood is a shred of essential evidence for reconstructing crime scenes. Blood identification and classification may help to confirm a suspect, although several chemical processes are used to recreate the crime scene. However, these approaches can have an impact on DNA analysis. A potential application of bloodstain identification and classification using Hyperspectral Imaging (HSI) can be used as substance clas-sification in forensic science for crime scene analysis. Therefore, this work proposes the use of a fast and compact Hybrid CNN to process HSI data for bloodstain identification and classification. For experimental and validation purposes, we perform exper-iments on a publicly available Hyperspectral-based Bloodstain dataset. This dataset has different types of substances i.e., blood and blood-like compounds, for instance, ketchup, artificial blood, beetroot juice, poster paint, tomato concentrate, acrylic paint, uncertain blood. We compare the results with state-of-the-art 3D CNN model and examine the results in detail and present a discussion of each tested architecture with limited availability of the training samples (e.g., only 5 % (792 samples) of the data samples are used to train the model, and validated on 5 % (792 samples) data samples and finally blindly tested on 90 % (14260 samples) of the data samples). The source code can be access on https://github.com/MHassaanButt/FCHCNN-for-HSIC
{"title":"A Fast and Compact Hybrid CNN for Hyperspectral Imaging-based Bloodstain Classification","authors":"Muhammad Hassaan Farooq Butt, Hamail Ayaz, Muhammad Ahmad, J. Li, R. Kuleev","doi":"10.1109/CEC55065.2022.9870277","DOIUrl":"https://doi.org/10.1109/CEC55065.2022.9870277","url":null,"abstract":"In forensic sciences, blood is a shred of essential evidence for reconstructing crime scenes. Blood identification and classification may help to confirm a suspect, although several chemical processes are used to recreate the crime scene. However, these approaches can have an impact on DNA analysis. A potential application of bloodstain identification and classification using Hyperspectral Imaging (HSI) can be used as substance clas-sification in forensic science for crime scene analysis. Therefore, this work proposes the use of a fast and compact Hybrid CNN to process HSI data for bloodstain identification and classification. For experimental and validation purposes, we perform exper-iments on a publicly available Hyperspectral-based Bloodstain dataset. This dataset has different types of substances i.e., blood and blood-like compounds, for instance, ketchup, artificial blood, beetroot juice, poster paint, tomato concentrate, acrylic paint, uncertain blood. We compare the results with state-of-the-art 3D CNN model and examine the results in detail and present a discussion of each tested architecture with limited availability of the training samples (e.g., only 5 % (792 samples) of the data samples are used to train the model, and validated on 5 % (792 samples) data samples and finally blindly tested on 90 % (14260 samples) of the data samples). The source code can be access on https://github.com/MHassaanButt/FCHCNN-for-HSIC","PeriodicalId":153241,"journal":{"name":"2022 IEEE Congress on Evolutionary Computation (CEC)","volume":"33 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":"125644706","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/CEC55065.2022.9870414
Xinyang Du, Ruibin Bai, Tianxiang Cui, R. Qu, Jiawei Li
A competitive traveling salesmen problem is a variant of traveling salesman problem in that multiple agents compete with each other in visiting a number of cities. The agent who is the first one to visit a city will receive a reward. Each agent aims to collect as more rewards as possible with the minimum traveling distance. There is still not effective algorithms for this complicated decision making problem. We investigate an improved ant colony approach for the competitive traveling sales-men problem which adopts a time dominance mechanism and a revised pheromone depositing method to improve the quality of solutions with less computational complexity. Simulation results show that the proposed algorithm outperforms the state of art algorithms.
{"title":"An Improved Ant Colony Approach for the Competitive Traveling Salesmen Problem","authors":"Xinyang Du, Ruibin Bai, Tianxiang Cui, R. Qu, Jiawei Li","doi":"10.1109/CEC55065.2022.9870414","DOIUrl":"https://doi.org/10.1109/CEC55065.2022.9870414","url":null,"abstract":"A competitive traveling salesmen problem is a variant of traveling salesman problem in that multiple agents compete with each other in visiting a number of cities. The agent who is the first one to visit a city will receive a reward. Each agent aims to collect as more rewards as possible with the minimum traveling distance. There is still not effective algorithms for this complicated decision making problem. We investigate an improved ant colony approach for the competitive traveling sales-men problem which adopts a time dominance mechanism and a revised pheromone depositing method to improve the quality of solutions with less computational complexity. Simulation results show that the proposed algorithm outperforms the state of art algorithms.","PeriodicalId":153241,"journal":{"name":"2022 IEEE Congress on Evolutionary Computation (CEC)","volume":"7 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":"124658217","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/CEC55065.2022.9870341
Arisa Toda, S. Hiwa, Kensuke Tanioka, Tomoyuki Hiroyasu
Topological Data Analysis (TDA) is an analytical technique that can reveal the skeletal structure inherent in complex or high-dimensional data. In this study, we considered the optimization search trajectories obtained from multiple trials of evolutionary computation as a single data set and challenged to represent the similarities and differences of each search trajectory as a topological network. Mapper is one of TDA tools and it includes the dimensionality reduction of data and clustering during graph generation. We modified Mapper to apply into this problem. The proposed framework is Mapper for evolutionary computation (EvoMapper). In the numerical experiments, multiple searches were conducted at different initial points to provide a basic review of the effectiveness of EvoMapper. The test functions were the One-max and Rastrigin function. A graph providing intuitive insights on the analysis results was constructed and visualized. In addition, the trials that reached the optimal solution and those that did not were clustered and found to have similar topology.
{"title":"Visualization, Clustering, and Graph Generation of Optimization Search Trajectories for Evolutionary Computation Through Topological Data Analysis: Application of the Mapper","authors":"Arisa Toda, S. Hiwa, Kensuke Tanioka, Tomoyuki Hiroyasu","doi":"10.1109/CEC55065.2022.9870341","DOIUrl":"https://doi.org/10.1109/CEC55065.2022.9870341","url":null,"abstract":"Topological Data Analysis (TDA) is an analytical technique that can reveal the skeletal structure inherent in complex or high-dimensional data. In this study, we considered the optimization search trajectories obtained from multiple trials of evolutionary computation as a single data set and challenged to represent the similarities and differences of each search trajectory as a topological network. Mapper is one of TDA tools and it includes the dimensionality reduction of data and clustering during graph generation. We modified Mapper to apply into this problem. The proposed framework is Mapper for evolutionary computation (EvoMapper). In the numerical experiments, multiple searches were conducted at different initial points to provide a basic review of the effectiveness of EvoMapper. The test functions were the One-max and Rastrigin function. A graph providing intuitive insights on the analysis results was constructed and visualized. In addition, the trials that reached the optimal solution and those that did not were clustered and found to have similar topology.","PeriodicalId":153241,"journal":{"name":"2022 IEEE Congress on Evolutionary Computation (CEC)","volume":"257 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":"134100853","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/CEC55065.2022.9870303
J. C. Garbelini, D. Sanches, A. Pozo
Finding transcription factor binding sites plays an important role inside bioinformatics. Its correct identification in the promoter regions of co-expressed genes is a crucial step for understanding gene expression mechanisms and creating new drugs and vaccines. The problem of finding motifs consists in seeking conserved patterns in biological datasets of sequences, through using unsupervised learning algorithms. This problem is considered one of the open problems of computational biology, which in its simplest formulation has been proven to be np-hard. Moreover, heuristics and meta-heuristics algorithms have been shown to be very promising in solving combinatorial problems with very large search spaces. In this paper we propose a new algorithm called Biomapp (Biological Motif Application) based on canonical Expectation Maximization that uses the Kullback-Leibler divergence to re-estimate the parameters of statistical model. Furthermore, the algorithm is embedded in an Iterated Local Search, as the local search step and then, we use a hierarchical perturbation operator in order to escape from local optima. The results obtained by this new approach were compared with the state-of-the-art algorithm MEME (Multiple EM Motif Elicitation) showing that Biomapp outperformed this classical technique in several datasets.
寻找转录因子结合位点在生物信息学中起着重要的作用。在共表达基因的启动子区域正确识别它是理解基因表达机制和创造新药和疫苗的关键一步。寻找基序的问题在于通过使用无监督学习算法在序列的生物数据集中寻找保守模式。这个问题被认为是计算生物学的开放问题之一，其最简单的表述已被证明是np困难的。此外，启发式和元启发式算法已被证明在解决具有非常大搜索空间的组合问题方面非常有前途。本文提出了一种基于典型期望最大化的新算法Biomapp (Biological Motif Application)，该算法利用Kullback-Leibler散度对统计模型的参数进行重新估计。此外，将算法嵌入到迭代局部搜索中，作为局部搜索步骤，然后使用层次摄动算子来避免局部最优。通过这种新方法获得的结果与最先进的算法MEME (Multiple EM Motif Elicitation)进行了比较，表明Biomapp在几个数据集中优于这种经典技术。
{"title":"Expectation Maximization based algorithm applied to DNA sequence motif finder","authors":"J. C. Garbelini, D. Sanches, A. Pozo","doi":"10.1109/CEC55065.2022.9870303","DOIUrl":"https://doi.org/10.1109/CEC55065.2022.9870303","url":null,"abstract":"Finding transcription factor binding sites plays an important role inside bioinformatics. Its correct identification in the promoter regions of co-expressed genes is a crucial step for understanding gene expression mechanisms and creating new drugs and vaccines. The problem of finding motifs consists in seeking conserved patterns in biological datasets of sequences, through using unsupervised learning algorithms. This problem is considered one of the open problems of computational biology, which in its simplest formulation has been proven to be np-hard. Moreover, heuristics and meta-heuristics algorithms have been shown to be very promising in solving combinatorial problems with very large search spaces. In this paper we propose a new algorithm called Biomapp (Biological Motif Application) based on canonical Expectation Maximization that uses the Kullback-Leibler divergence to re-estimate the parameters of statistical model. Furthermore, the algorithm is embedded in an Iterated Local Search, as the local search step and then, we use a hierarchical perturbation operator in order to escape from local optima. The results obtained by this new approach were compared with the state-of-the-art algorithm MEME (Multiple EM Motif Elicitation) showing that Biomapp outperformed this classical technique in several datasets.","PeriodicalId":153241,"journal":{"name":"2022 IEEE Congress on Evolutionary Computation (CEC)","volume":"138 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":"131795880","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/CEC55065.2022.9870396
Yi-Feng Chang, Chuan-Kang Ting
Floorplanning is a crucial process in the early stage of VLSI physical design. It determines the performance, reliability, and size of chips. B*-tree is a simple yet efficient representation that encodes the layout of modules in a compact and non-slicing structure. Several B*-tree variants and corresponding operators have been proposed to deal with non-slicing floorplanning. However, these operators are considered and applied individually. A collective manipulation of them remains missing. This study proposes a genetic algorithm (GA) that enables multiple crossover and mutation operators for solving the non-slicing floorplanning problem. In particular, the GA selects one crossover operator and one mutation operator from the pool of operators whenever reproducing an offspring. The probability for an operator to be selected is based on its empirical performance. This study conducts experiments on two well-known benchmarks to examine the effectiveness of the proposed method. The experimental results show that the GA can achieve superior solution quality and efficiency on the non-slicing VLSI floorplanning.
{"title":"Multiple Crossover and Mutation Operators Enabled Genetic Algorithm for Non-slicing VLSI Floorplanning","authors":"Yi-Feng Chang, Chuan-Kang Ting","doi":"10.1109/CEC55065.2022.9870396","DOIUrl":"https://doi.org/10.1109/CEC55065.2022.9870396","url":null,"abstract":"Floorplanning is a crucial process in the early stage of VLSI physical design. It determines the performance, reliability, and size of chips. B*-tree is a simple yet efficient representation that encodes the layout of modules in a compact and non-slicing structure. Several B*-tree variants and corresponding operators have been proposed to deal with non-slicing floorplanning. However, these operators are considered and applied individually. A collective manipulation of them remains missing. This study proposes a genetic algorithm (GA) that enables multiple crossover and mutation operators for solving the non-slicing floorplanning problem. In particular, the GA selects one crossover operator and one mutation operator from the pool of operators whenever reproducing an offspring. The probability for an operator to be selected is based on its empirical performance. This study conducts experiments on two well-known benchmarks to examine the effectiveness of the proposed method. The experimental results show that the GA can achieve superior solution quality and efficiency on the non-slicing VLSI floorplanning.","PeriodicalId":153241,"journal":{"name":"2022 IEEE Congress on Evolutionary Computation (CEC)","volume":"141 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":"133534282","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/CEC55065.2022.9870219
Julian Blank, K. Deb
Metaheuristics are known to be effective for solving a broad category of optimization problems. However, most heuristics require different parameter settings appropriately for a problem class or even for a specific problem. Researchers address this commonly by performing a parameter tuning study (also known as hyper-parameter optimization) or developing a parameter control mechanism that changes parameters dynamically. Whereas parameter tuning is computationally expensive and limits the parameter configuration to stay constant throughout the run, parameter control is also a challenging task because all dynamics induced by various operators must be learned to make an appropriate adaptation of parameters on the fly. This paper investigates parameter tuning and control for a well-known optimization method - differential evolution (DE). In contrast to most existing DE practices, an additional individualistic evolutionary operator called polynomial mutation is incorporated into the offspring creation. Results on test problems with up to 50 variables indicate that mutation can be helpful for multi-modal problems to escape from local optima. On the one hand, the effectiveness of parameter tuning for a specific problem becomes apparent; on the other hand, its generalization capabilities seem to be limited. Moreover, a generic coevolutionary approach for parameter control outperforms a random choice of parameters. Recognizing the importance of choosing a suitable parameter configuration to solve any optimization problem, we have incorporated a standard implementation of both tuning and control approaches into a single framework, providing a direction for the evolutionary computation and optimization researchers to use and further investigate the effects of parameters on DE and other metaheuristics-based algorithms.
{"title":"Parameter Tuning and Control: A Case Study on Differential Evolution With Polynomial Mutation","authors":"Julian Blank, K. Deb","doi":"10.1109/CEC55065.2022.9870219","DOIUrl":"https://doi.org/10.1109/CEC55065.2022.9870219","url":null,"abstract":"Metaheuristics are known to be effective for solving a broad category of optimization problems. However, most heuristics require different parameter settings appropriately for a problem class or even for a specific problem. Researchers address this commonly by performing a parameter tuning study (also known as hyper-parameter optimization) or developing a parameter control mechanism that changes parameters dynamically. Whereas parameter tuning is computationally expensive and limits the parameter configuration to stay constant throughout the run, parameter control is also a challenging task because all dynamics induced by various operators must be learned to make an appropriate adaptation of parameters on the fly. This paper investigates parameter tuning and control for a well-known optimization method - differential evolution (DE). In contrast to most existing DE practices, an additional individualistic evolutionary operator called polynomial mutation is incorporated into the offspring creation. Results on test problems with up to 50 variables indicate that mutation can be helpful for multi-modal problems to escape from local optima. On the one hand, the effectiveness of parameter tuning for a specific problem becomes apparent; on the other hand, its generalization capabilities seem to be limited. Moreover, a generic coevolutionary approach for parameter control outperforms a random choice of parameters. Recognizing the importance of choosing a suitable parameter configuration to solve any optimization problem, we have incorporated a standard implementation of both tuning and control approaches into a single framework, providing a direction for the evolutionary computation and optimization researchers to use and further investigate the effects of parameters on DE and other metaheuristics-based algorithms.","PeriodicalId":153241,"journal":{"name":"2022 IEEE Congress on Evolutionary Computation (CEC)","volume":"24 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":"121613872","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/CEC55065.2022.9870207
Kohei Fukuhara, Ryo Kumagai, Fukawa Yuta, S. Tanabe, Hiroki Kawano, Yoshihiro Ohta, Hiroyuki Sato
This work addresses a real-world building facility control problem by using evolutionary algorithms. The variables are facility control parameters, such as the start/stop time of air-conditioning, lighting, and ventilation operation, etc. The problem has six objectives: annual energy consumption, elec-tricity cost, overall satisfaction, thermal satisfaction, indoor air quality satisfaction, and lighting satisfaction. The problem has five constraints: power consumption, temperature, humidity, $mathbf{CO}_{2}$ concentration, and average illuminance. To solve the problem, we utilize IBEA framework. For efficient solution generation, we employ the steady-state model for IBEA. We propose the total constraint win-loss rank for multiple constraints to treat multiple constraints equally. Experimental results on artificial test problems and building facility control problems show that the proposed constraint IBEA with steady-state and total con-straint win-loss rank archives better search performance than conventional representative algorithms.
{"title":"Digital Twin Based Evolutionary Building Facility Control Optimization","authors":"Kohei Fukuhara, Ryo Kumagai, Fukawa Yuta, S. Tanabe, Hiroki Kawano, Yoshihiro Ohta, Hiroyuki Sato","doi":"10.1109/CEC55065.2022.9870207","DOIUrl":"https://doi.org/10.1109/CEC55065.2022.9870207","url":null,"abstract":"This work addresses a real-world building facility control problem by using evolutionary algorithms. The variables are facility control parameters, such as the start/stop time of air-conditioning, lighting, and ventilation operation, etc. The problem has six objectives: annual energy consumption, elec-tricity cost, overall satisfaction, thermal satisfaction, indoor air quality satisfaction, and lighting satisfaction. The problem has five constraints: power consumption, temperature, humidity, $mathbf{CO}_{2}$ concentration, and average illuminance. To solve the problem, we utilize IBEA framework. For efficient solution generation, we employ the steady-state model for IBEA. We propose the total constraint win-loss rank for multiple constraints to treat multiple constraints equally. Experimental results on artificial test problems and building facility control problems show that the proposed constraint IBEA with steady-state and total con-straint win-loss rank archives better search performance than conventional representative algorithms.","PeriodicalId":153241,"journal":{"name":"2022 IEEE Congress on Evolutionary Computation (CEC)","volume":"10 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":"117309164","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/CEC55065.2022.9870315
Jonathan Carrero, Ismael Rodríguez, F. Rubio
In multi-agent resource allocation systems, it is reasonable that the specific allocation of resources depends on the utility functions declared by the different agents. However, this can easily lead to strategic behaviors in which the agents involved are interested in lying, since such lies can bring them more profitable deals. In this paper we analyze the case of egalitarian social welfare, where the objective is to maximize the utility of the agent who receives the least utility. In this context, agents can obtain advantages by undervaluing their preferences. Thus, we will see how to discourage such lies even in the presence of public goods and non-additive utilities. Likewise, we will use genetic algorithms to show, through experimental results, the robustness of our proposal against lies.
{"title":"Avoiding strategic behaviors in the egalitarian social welfare under public resources and non-additive utilities","authors":"Jonathan Carrero, Ismael Rodríguez, F. Rubio","doi":"10.1109/CEC55065.2022.9870315","DOIUrl":"https://doi.org/10.1109/CEC55065.2022.9870315","url":null,"abstract":"In multi-agent resource allocation systems, it is reasonable that the specific allocation of resources depends on the utility functions declared by the different agents. However, this can easily lead to strategic behaviors in which the agents involved are interested in lying, since such lies can bring them more profitable deals. In this paper we analyze the case of egalitarian social welfare, where the objective is to maximize the utility of the agent who receives the least utility. In this context, agents can obtain advantages by undervaluing their preferences. Thus, we will see how to discourage such lies even in the presence of public goods and non-additive utilities. Likewise, we will use genetic algorithms to show, through experimental results, the robustness of our proposal against lies.","PeriodicalId":153241,"journal":{"name":"2022 IEEE Congress on Evolutionary Computation (CEC)","volume":"102 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":"117275064","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}
Multimodal optimization problems (MMOPs) are sorts of optimization problems that have many global optima. To discover as many peaks as possible and increase the accuracy of the solutions, MMOP requires algorithms with great exploration and exploitation abilities. However, exploration and exploitation are in an adversarial relationship, since exploration aims to locate more optima via searching the global space rather than small regions, whereas exploitation targets to enhance the accuracy of solutions via searching in small areas. The key to efficiently solving MMOPs lies in striking a balance between exploration and exploitation. To achieve the goal, this paper proposes an adversarial differential evolution (ADE), containing an adversarial reproduction strategy and an adversarial selection strategy. Firstly, adversarial reproduction strategy generates offspring for exploration and offspring for exploitation and lets these two types of offspring compete for survival. Secondly, adversarial selection strategy employs a diversity-optimization-based selection and a crowding-based selection to select the offspring with both good diversity and good fitness. Diversity-optimization-based selection transforms the problem of selecting diverse individuals into an optimizing problem and solves it via an extra genetic algorithm to get the offspring with optimal diversity. Extensive experiments are conducted on CEC2013 MMOP benchmark to verify the effectiveness and efficiency of the proposed ADE. Experimental results show that ADE has advantages over the state-of-the-art MMOP algorithms.
{"title":"Adversarial Differential Evolution for Multimodal Optimization Problems","authors":"Yiyi Jiang, Chun-Hua Chen, Zhi-hui Zhan, Yunjie Li, Jinchao Zhang","doi":"10.1109/CEC55065.2022.9870298","DOIUrl":"https://doi.org/10.1109/CEC55065.2022.9870298","url":null,"abstract":"Multimodal optimization problems (MMOPs) are sorts of optimization problems that have many global optima. To discover as many peaks as possible and increase the accuracy of the solutions, MMOP requires algorithms with great exploration and exploitation abilities. However, exploration and exploitation are in an adversarial relationship, since exploration aims to locate more optima via searching the global space rather than small regions, whereas exploitation targets to enhance the accuracy of solutions via searching in small areas. The key to efficiently solving MMOPs lies in striking a balance between exploration and exploitation. To achieve the goal, this paper proposes an adversarial differential evolution (ADE), containing an adversarial reproduction strategy and an adversarial selection strategy. Firstly, adversarial reproduction strategy generates offspring for exploration and offspring for exploitation and lets these two types of offspring compete for survival. Secondly, adversarial selection strategy employs a diversity-optimization-based selection and a crowding-based selection to select the offspring with both good diversity and good fitness. Diversity-optimization-based selection transforms the problem of selecting diverse individuals into an optimizing problem and solves it via an extra genetic algorithm to get the offspring with optimal diversity. Extensive experiments are conducted on CEC2013 MMOP benchmark to verify the effectiveness and efficiency of the proposed ADE. Experimental results show that ADE has advantages over the state-of-the-art MMOP algorithms.","PeriodicalId":153241,"journal":{"name":"2022 IEEE Congress on Evolutionary Computation (CEC)","volume":"163 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":"116372597","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：