检索结果-内蒙古大学图书馆

Density peaks clustering algorithm based on multi-cluster merge and its application in the extraction of typical load patterns of users

引用

Journal of Ambient Intelligence and Humanized computing 2024年第8期15卷 3193-3209页

作者： Zhao, Jia Yao, Zhanfeng Qiu, Liujun Fan, Tanghuai Lee, Ivan School of Information Engineering Nanchang Institute of Technology Nanchang330099 China Nanchang Key Laboratory of IoT Perception and Collaborative Computing for Smart City Nanchang Institute of Technology Nanchang330099 China UniSA STEM University of South Australia AdelaideSA5000 Australia

The density peaks clustering (DPC) algorithm is simple in principle, efficient in operation, and has good clustering effects on various types of datasets. However, this algorithm still has some defects: (1) due to the definition limitations of local density and relative distance of samples, it is difficult for the algorithm to find correct density peaks;(2) the allocation strategy of the algorithm has poor robustness and is prone to cause other problems. In response to solve the above shortcomings, we proposed a density peaks clustering algorithm based on multi-cluster merge (DPC-MM). In view of the difficulty in selecting density peaks of the DPC algorithm, a new method of calculating relative distance of samples was defined to make the density peaks found more accurate. The allocation strategy of multi-cluster merge was proposed to alleviate or avoid problems caused by allocation errors. Experimental results revealed that the DPC-MM algorithm can efficiently perform clustering on datasets of any shape and scale. The DPC-MM algorithm was applied in extraction of typical load patterns of users, and can more accurately perform clustering on user loads. The extraction results can better reflect electricity consumption habits of users. © The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature 2024.

关键词： Clustering algorithms

来源：评论

学校读者我要写书评

暂无评论

Dual-Population-Iterating and Preys-Gathering MPA for XGBoost Optimization 16th

Dual-Population-Iterating and Preys-Gathering MPA for XGBoos...

引用

16th International Conference on Genetic and Evolutionary computing, ICGEC 2024

作者： Li, Jiacheng Lv, Li Kang, Ping Tian, Xiumei School of Information Engineering Nanchang Institute of Technology Nanchang330099 China Nanchang Key Laboratory of IoT Perception and Collaborative Computing for Smart City Nanchang Institute of Technology Nanchang330099 China

ISBN: (纸本)9789819615308

The marine predator algorithm (MPA) often struggles with issues such as low population diversity, susceptibility to local optima, and population aggregation when optimizing XGBoost, which is sensitive to numerous parameters. To address these challenges, we propose the dual-population-iterating and prey-gathering marine predator algorithm (DPGMPA) and apply it to XGBoost parameter optimization. Initially, we utilize a dual-population-iterating strategy to create the original population and an opposition-based learning population. Brownian motion and Laplace distribution respectively control the step size to guide population updates toward the global optimum, enhancing population diversity. Subsequently, the preys-gathering strategy is employed, selecting prey with the highest accuracy in each iteration and having neighboring preys converge towards them to mitigate local optima. Furthermore, we optimize the Fish Aggregating Devices (FADs) by introducing variable weight control, addressing the population aggregation issue. To validate the parameter optimization capability of this algorithm for XGBoost, we conducted experiments on six datasets, using accuracy and F1 score as evaluation metrics, and compared the results with five other algorithms. The experimental results demonstrate that the DPGMPA algorithm more effectively escapes local optima when optimizing the XGBoost classifier, significantly improving XGBoost's classification performance. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Laplace distribution Marine Predator Algorithm normal distribution parameter optimization population diversity XGBoost

来源：评论

学校读者我要写书评

暂无评论

Feature Intracorrelation and Weighted Mahalanobis Distance for Fuzzy Support Vector Machine 16th

Feature Intracorrelation and Weighted Mahalanobis Distance f...

引用

16th International Conference on Genetic and Evolutionary computing, ICGEC 2024

作者： Jiang, Lan Lv, Li Peng, Siwei Xie, Haihua School of Information Engineering Nanchang Institute of Technology Nanchang330099 China Nanchang Key Laboratory of IoT Perception and Collaborative Computing for Smart City Nanchang Institute of Technology Nanchang330099 China

ISBN: (纸本)9789819615308

Most feature-weighted fuzzy support vector machines only calculate the weight of each feature without considering inter-feature correlations, leading to redundant features that significantly impact final classification results. Additionally, designing fuzzy affiliation based on Euclidean distance overlooks the overall sample distribution and inter-feature coupling. To address these issues, we propose a fuzzy support vector machine incorporating intra-feature correlation and weighted Mahalanobis distance. First, ReliefF is employed to calculate feature weights, while the Pearson algorithm introduces a penalty factor based on the Pearson coefficient between features, updating the weights to mitigate the impact of redundant features on classification performance. Second, Mahalanobis distance replaces Euclidean distance, and a feature-weighted subordinate function based on Mahalanobis distance is constructed using the covariance matrix to eliminate variance differences between dimensions, ensuring feature independence and uniform distribution. Finally, a feature-weighted kernel function is integrated to prevent weakly correlated or uncorrelated features from dominating kernel function calculations. Accuracy and F1 value are chosen as evaluation metrics and compared with existing algorithms on the UCI dataset. Experimental results demonstrate that this algorithm effectively reduces the interference of weakly correlated, irrelevant, and redundant features while considering sample distribution, leading to superior classification performance. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Support vector machines

来源：评论

学校读者我要写书评

暂无评论

Tolerance-based Elite Mutation Firefly Algorithm

Journal of Network Intelligence

引用

Journal of Network Intelligence 2023年第4期8卷 1406-1425页

作者： Chen, Juan Zeng, Ren-Xian Qiu, Shen-Yu Kang, Ping Zhao, Jia School of Information Engineering Nanchang Institute of Technology Nanchang Key Laboratory of IoT Perception and Collaborative Computing for Smart City China School of Information Engineering Nanchang Institute of Technology China College of Science Nanchang Institute of Technology China

For the firefly algorithm in the solution of complex functions are prone to low accuracy and "premature" phenomenon. In this paper, we propose a tolerance-based elite mutation firefly algorithm (MOFA-TEM). The algorithm employs non-dominated sorting and congestion distance to select the elite leader, which is fused with the individual historical optimal solution to form a more comprehensive integrated information of the population, and jointly guide the movement of the fireflies, which not only en-hances the interactivity of the information between fireflies in the population, but also improves the algorithm’s ability of global exploration. Secondly, a tolerance-based firefly state judgment mechanism is designed, which will exponentially increase the probability of updating the elite leader as the number of individual firefly stagnation increases, which both fully exploits the potential of ’s search for excellence and prevents ineffective search-ing of the population. Finally, if the firefly triggers the tolerance mechanism, the elite mutation strategy will be used to update the and select a new elite learning object to avoid the algorithm from stagnating. Experiments on 19 test functions and validation using Friedman’s test show that MOFA-TEM gives better results compared to other optimization algorithms. © 2023.

关键词： Bioluminescence

来源：评论

学校读者我要写书评

暂无评论

Double Replay Buffer Mechanism for Deep Reinforcement Learning Algorithm 4

Double Replay Buffer Mechanism for Deep Reinforcement Learni...

引用

4th International Conference on Electronic Information Engineering and Computer Science, EIECS 2024

作者： Ji, Yiwei Le, Weixuan Tian, Xiumei Liu, Guangming Han, Longzhe Cui, Tao Nanchang Institute of Technology Nanchang Key Laboratory of IoT Perception and Collaborative Computing for Smart City Nanchang China School of Electrical and Electronic Engineering Changchun University of Technology Changchun China

ISBN: (纸本)9798331531409

Deep Reinforcement Learning (DRL) has emerged as a highly promising area of research within the fields of artificial intelligence and machine learning. However, the conventional experience replay mechanism employed in DRL typically utilizes a uniform sampling method, which fails to account for the varying importance of different samples. This oversight can result in suboptimal sample utilization and low learning efficiency. To address these limitations, this paper proposes a double experience replay buffer mechanism. The proposed approach adaptively adjusts the experience pool ratio through the Q-learning algorithm and leverages the Upper Confidence Bound (UCB) method to calculate sample priorities. Furthermore, we design a hybrid sampling strategy to increase the selection rate of high-value samples, and enhance overall learning performance. To validate the effectiveness of this approach, comprehensive experiments were conducted across several Gym control tasks. The experimental results demonstrate the improvements in learning efficiency and performance. © 2024 IEEE.

关键词： Deep reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

An attention based proximal policy optimization algorithm for imitation reinforcement learning

An attention based proximal policy optimization algorithm fo...

引用

2024 International Conference on Computer Graphics, Artificial Intelligence, and Data Processing, ICCAID 2024

作者： Zhang, Hui Xia, Yu Chen, Li Hu, Rong Han, Longzhe Nanchang Key Laboratory of IoT Perception and Collaborative Computing for Smart City Nanchang Institute of Technology 289 Tianxiang Avenue Nanchang330099 China

ISBN: (纸本)9781510689275

Deep Reinforcement Learning (DRL) has shown great potential in addressing complex decision-making challenges, especially within high-dimensional and dynamic environments. However, DRL faces limitations, such as low sample efficiency and difficulties in reward function design, which restrict its practical applicability. To address these challenges, the paper presents an Attention-based Proximal Policy Optimization Algorithm for Imitation Reinforcement Learning (APPO-IRL). APPO-IRL leverages Squeeze and Excitation modules to compress channel features, enabling efficient computation of importance weights with minimal computational overhead. The approach significantly enhances Behavioral Cloning’s (BC) ability to capture critical features. Additionally, a deep ensemble method is integrated within BC to reduce uncertainty when generating expert policies. The policy update is then guided by the Kullback-Leibler divergence between expert and RL policies. Experimental results demonstrate that APPO-IRL outperforms traditional DRL algorithms by reducing the need for complex reward function design and significantly improving sample efficiency and overall performance. Compared to BC-based DRL approaches, APPO-IRL also shows marked improvements in stability and convergence. © 2025 SPIE.

关键词： Deep reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

Double Replay Buffer Mechanism for Deep Reinforcement Learning Algorithm

Double Replay Buffer Mechanism for Deep Reinforcement Learni...

引用

Electronic Information Engineering and Computer Science (EIECS), 2021 International Conference on

作者： Yiwei Ji Weixuan Le Xiumei Tian Guangming Liu Longzhe Han Tao Cui Nanchang Key Laboratory of IoT Perception and Collaborative Computing for Smart City Nanchang Institute of Technology Nanchang China School of Electrical and Electronic Engineering Changchun University of Technology Changchun China

ISBN: (数字)9798331531409

ISBN: (纸本)9798331531416

关键词： Computer science Q-learning Upper bound Deep reinforcement learning Sampling methods

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：