检索结果-内蒙古大学图书馆

Deep Deterministic Policy Gradient With Compatible Critic Network

IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS 2023年第8期34卷 4332-4344页

作者： Wang, Di Hu, Mengqi Univ Illinois Dept Mech & Ind Engn Chicago IL 60607 USA

Deep deterministic policy gradient (DDPG) is a powerful reinforcement learning algorithm for large-scale continuous controls. DDPG runs the back-propagation from the state-action value function to the actor network's parameters directly, which raises a big challenge for the compatibility of the critic network. This compatibility emphasizes that the policy evaluation is compatible with the policy improvement. As proved in deterministic policy gradient, the compatible function guarantees the convergence ability but restricts the form of the critic network tightly. The complexities and limitations of the compatible function impede its development in DDPG. This article introduces neural networks' similarity indices with gradients to measure the compatibility concretely. Represented as kernel matrices, we consider the actor network's and the critic network's training dataset, trained parameters, and gradients. With the sketching trick, the calculation time of the similarity index decreases hugely. The centered kernel alignment index and the normalized Bures similarity index provide us with consistent compatibility scores empirically. Moreover, we demonstrate the necessity of the compatible critic network in DDPG from three aspects: 1) analyzing the policy improvement/evaluation steps;2) conducting the theoretic analysis;and 3) showing the experimental results. Following our research, we remodel the compatible function with an energy function model, enabling it suitable to the sizeable state-action space problem. The critic network has higher compatibility scores and better performance by introducing the policy change information into the critic-network optimization process. Besides, based on our experiment observations, we propose a light-computation overestimation solution. To prove our algorithm's performance and validate the compatibility of the critic network, we compare our algorithm with six state-of-the-art algorithms using seven PyBullet robotics environments.

关键词： Training Kernel Indexes approximation algorithms Neural networks Estimation Convergence Compatible function deep deterministic policy gradient (DDPG) overestimation reinforcement learning (RL)

来源：评论

学校读者我要写书评

暂无评论

Cost-Sensitive Online Adaptive Kernel Learning for Large-Scale Imbalanced Classification

引用

IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING 2023年第10期35卷 10554-10568页

作者： Chen, Yingying Hong, Zijie Yang, Xiaowei South China Univ Technol Sch Software Engn Guangzhou 510006 Peoples R China

Imbalanced classification is a challenging task in the fields of machine learning, data mining and pattern recognition. Cost-sensitive online algorithms are very important methods for large-scale imbalanced classification problems. At present, most of the cost-sensitive classification algorithms focus on the accuracy of the minority class and ignore the accuracy of the majority class. In order to better balance the accuracy between the minority class and the majority class, in this article, a misclassification cost is presented to ensure that the cost-sensitive online algorithm can better deal with the imbalanced classification problems without signifificantly reducing the accuracy of the majority class. Based on the proposed misclassification cost, a novel cost-sensitive online adaptive kernel learning algorithm is proposed to boost the adaptability of kernel function when data arrives one by one. According to the essential characteristics of the imbalanced binary classification, a cost-sensitive online adaptive kernel learning algorithm is given to handle the large-scale imbalanced multi-class classification problems. Theoretical analysis of the proposed algorithms are provided. Extensive experiments demonstrate that compared with the state-of-the-art imbalanced classification algorithms, the proposed algorithms can significantly improve the classification performances on most of the large-scale imbalanced data sets.

关键词： Classification algorithms Kernel Costs Adaptation models Machine learning algorithms approximation algorithms Task analysis Adaptive algorithms cost function classification algorithms kernel machine learning algorithms

来源：评论

学校读者我要写书评

暂无评论

Finite-Time Adaptive Fuzzy Prescribed Performance Formation Control for High-Order Nonlinear Multiagent Systems Based on Event-Triggered Mechanism

引用

IEEE TRANSACTIONS ON FUZZY SYSTEMS 2023年第4期31卷 1229-1239页

作者： Zhou, Haodong Sui, Shuai Tong, Shaocheng Liaoning Univ Technol Coll Sci Jinzhou 121001 Peoples R China

This article investigates the finite-time adaptive fuzzy prescribed performance formation control problem for high-order nonlinear multiagent systems. Fuzzy logic systems (FLSs) are adopted to approximate the unknown nonlinear dynamics. By combining the dynamic surface control technique and backstepping recursive design, an adaptive fuzzy prescribed performance formation control algorithm is developed. To reduce unnecessary transmission of network resources, a novel event-triggered mechanism is proposed in the control method. Subsequently, based on the adding power integral method and finite-time stability theory, it is proved that all signals of the controlled system are bounded and the tracking errors do not exceed the prescribed performance bounds in a finite time. Simulation results are given to verify that the presented formation control scheme achieves desired results.

关键词： Formation control Fuzzy logic Control systems Artificial neural networks Heuristic algorithms Backstepping approximation algorithms Event-triggered mechanism finite-time forma- tion control high-order nonlinear multiagent systems (MASs) prescribed performance

来源：评论

学校读者我要写书评

暂无评论

Direct-Optimization-Based DC Dictionary Learning With the MCP Regularizer

引用

IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS 2023年第7期34卷 3568-3579页

作者： Li, Zhenni Yang, Zuyuan Zhao, Haoli Xie, Shengli Guangdong Univ Technol Sch Automat Guangzhou 510006 Peoples R China Guangdong HongKong Macao Joint Lab Smart Discrete Guangzhou 510006 Peoples R China Guangdong Key Lab IoT Informat Technol GDUT Guangzhou 510006 Peoples R China Sun Yat Sen Univ Sch Data & Comp Sci Guangzhou 510006 Peoples R China Minist Educ Key Lab Intelligent Informat Proc & Syst Integrat Guangzhou 510006 Peoples R China

Direct-optimization-based dictionary learning has attracted increasing attention for improving computational efficiency. However, the existing direct optimization scheme can only be applied to limited dictionary learning problems, and it remains an open problem to prove that the whole sequence obtained by the algorithm converges to a critical point of the objective function. In this article, we propose a novel direct-optimization-based dictionary learning algorithm using the minimax concave penalty (MCP) as a sparsity regularizer that can enforce strong sparsity and obtain accurate estimation. For solving the corresponding optimization problem, we first decompose the nonconvex MCP into two convex components. Then, we employ the difference of the convex functions algorithm and the nonconvex proximal-splitting algorithm to process the resulting subproblems. Thus, the direct optimization approach can be extended to a broader class of dictionary learning problems, even if the sparsity regularizer is nonconvex. In addition, the convergence guarantee for the proposed algorithm can be theoretically proven. Our numerical simulations demonstrate that the proposed algorithm has good convergence performances in different cases and robust dictionary-recovery capabilities. When applied to sparse approximations, the proposed approach can obtain sparser and less error estimation than the different sparsity regularizers in existing methods. In addition, the proposed algorithm has robustness in image denoising and key-frame extraction.

关键词： Machine learning Optimization Dictionaries Convergence Convex functions Signal processing algorithms approximation algorithms Convergence analysis dictionary learning direct optimization minimax concave penalty (MCP) regularizer

来源：评论

学校读者我要写书评

暂无评论

Covering Users With QoS by a Connected Swarm of Drones: Graph Theoretical Approach and Experiments

引用

IEEE-ACM TRANSACTIONS ON NETWORKING 2023年第6期31卷 2483-2498页

作者： Danilchenko, Kiril Nutov, Zeev Segal, Michael Ben Gurion Univ Negev Sch Elect & Comp Engn IL-84105 Beer Sheva Israel Open Univ Israel Comp Sci Dept IL-4353701 Raanana Israel

In this work, we study the connected version of the covering problem motivated by the coverage of ad-hoc drones' swarm. We focus on the situation where the number of drones is given, and this number is not necessarily enough to cover all users. That is, we deal with a budget optimization problem, where the budget is the number of given drones. We assume that each ground user has different QoS requirements. Additionally, each ground user has a weight that corresponds to the importance (rank) of the user. Moreover, we consider the case when there is no third-party entity that provides connectivity to the drones. In this paper, we propose a 3D deployment scheme with the given number of drones such that the sum of the weights (ranks) of the ground users covered by drones is maximized (when the covering radii satisfy QoS of these users), and the drones form a connected graph. We present a number of approximate solutions with provable guaranteed performance evaluation that have been validated also through the simulation platform.

关键词： Drones Quality of service Wireless communication Atmospheric modeling Signal to noise ratio approximation algorithms Three-dimensional displays Connected swarm of drones approximation algorithm coverage problem

来源：评论

学校读者我要写书评

暂无评论

Phase Retrieval Using Deep Dual Alternating Direction Method of Multipliers Network With Deep Sparse Prior Knowledge

引用

IEEE TRANSACTIONS ON AUDIO SPEECH AND LANGUAGE PROCESSING 2025年 33卷 557-569页

作者： Kim, Moogyeong Chung, Wonzoo Korea Univ Dept Artificial Intelligence Seoul 02841 South Korea

In this paper, a deep phase retrieval algorithm for speech signals based on the dual Alternating Direction Method of Multipliers (ADMM) incorporating a deep prior network that exploits the sparsity of speech signals is presented. The proposed network, named DADMM-net, unfolds the dual ADMM for the $\ell _{1}$-regularized non-convex optimization problem of phase retrieval with several two-dimensional convolutional neural networks (2D-CNNs). In order to efficiently optimize the deep unfolding network for high-dimensional parameter vectors, a novel updating scheme referred to as soft coordinate descent (soft-CD) is proposed, where dual parameter updates are determined through interpolation between the current values and the updated values coordinate-wise with respect to the weights computed by deep networks in each layer. Numerical simulations on a publicly available dataset confirm the state-of-the-art performance of the proposed method in terms of perceptual evaluation of speech quality and short-time objective intelligibility with a significantly faster convergence speed compared to existing methods.

关键词： Optimization Convex functions Time-domain analysis Spectrogram Signal processing algorithms Vectors Sparse approximation approximation algorithms Knowledge engineering Time-frequency analysis Deep sparse prior deep unfolding dual ADMM non-convex optimization phase retrieval

来源：评论

学校读者我要写书评

暂无评论

An Assessment of Deep Learning Versus Massively Parallel, Non-Linear Methods for Highly-Efficient MIMO Detection

引用

IEEE ACCESS 2023年 11卷 97493-97502页

作者： Ducoing, J. C. De Luna Jayawardena, Chathura Nikitopoulos, Konstantinos Univ Surrey Inst Commun Syst 5G & 6G Innovat Ctr Wireless Syst Lab Guildford GU2 7XH England

Multiple-user, multiple-input, multiple-output (MU-MIMO) systems supporting a large number of concurrent streams have the potential to substantially improve the connectivity and throughput of future wireless communication systems. Towards this goal, deep learning (DL)-based techniques have recently been proposed for MIMO signal detection. Good performance results have been reported when compared to conventional detection methods, but it is unclear how they measure against state-of-the-art detection techniques. In this work, for the first time, we perform a critical evaluation of DetNet, MMNet, GEPNet, and RE-MIMO, four prominent model-based DL techniques based on different working principles, and assess their reliability, complexity, and robustness against the practical Massively Parallel Non-Linear processing (MPNL) detection approach. The results show that the model-based DL approaches offer promising results but have difficulty adapting to channel models that differ from those on which they were trained. They also exhibit lower reliability and higher complexity than MPNL, even without considering the training stage. We find that, at present, the human-designed MPNL outperforms the DL-based detection methods in virtually all the metrics. Nevertheless, DL-based solutions are rapidly advancing, and further research intended to address their current shortcomings may one day offer advantages over human-designed detection methods.

关键词： MIMO communication Complexity theory Deep learning Adaptation models Wireless communication Signal detection approximation algorithms Parallel processing Multi-user MIMO (MU-MIMO) signal detection machine learning deep learning parallel processing

来源：评论

学校读者我要写书评

暂无评论

A Parallel Algorithm to Accelerate DEVS Simulations in Shared Memory Architectures

引用

IEEE TRANSACTIONS ON PARALLEL AND DISTRIBUTED SYSTEMS 2023年第5期34卷 1609-1620页

作者： Trabes, Guillermo German Wainer, Gabriel A. Gil-Costa, Veronica Carleton Univ Dept Syst & Comp Engn Ottawa ON K1S 5B6 Canada Univ Nacl San Luis RA-D5700 San Luis Argentina Natl Univ San Luis RA-D5700 San Luis Argentina CCT CONICET RA-D5700 San Luis Argentina

We propose a new algorithm for the execution of Discrete Event System Specification (DEVS) simulations on parallel shared memory architectures. Our approach executes parallel discrete-event simulations by executing all tasks in the PDEVS simulation protocol in parallel. The algorithm works by distributing the computations among different cores on shared memory architectures. To show the benefits of our algorithm, we present the results of a set of experiments using a synthetic benchmark and a real-world scenario using two independent computer architectures. The results obtained show how our algorithm accelerates simulations up to eight times, improving previous approaches. In addition, we show that our approach scales when we increase the number of CPU-cores used.

关键词： Computational modeling Protocols Synchronization Cadmium Task analysis approximation algorithms Software algorithms Discrete-event parallel algorithms simulation shared memory

来源：评论

学校读者我要写书评

暂无评论

Learning-Based Topology Optimization of Power Networks

引用

IEEE TRANSACTIONS ON POWER SYSTEMS 2023年第2期38卷 1366-1378页

作者： Han, Tong Hill, David J. Univ Hong Kong Dept Elect & Elect Engn Pokfulam Hong Kong Peoples R China Univ New South Wales Sch Elect Engn & Telecommun Sydney NSW 2052 Australia

Transitions of power networks demand more powerful solution approaches for optimal transmission switching (OTS) and distribution network reconfiguration (DNR) to tackle complicated cases with high computational efficiency. In this paper, we propose a learning-based solution approach for OTS and DNR, which is established based on an identified common pattern behind existing heuristic algorithms. The original hand-crafted and short-sighted evaluation function is substituted by a parameterized function whose architecture is well-designed by mainly exploiting the gated graph neural network and multilayer perceptrons. For learning the parameterized function that lacks training labels, we further design a learning algorithm that combines the double deep Q-network algorithm and multi-step learning. Then, with the learned parameterized function, the learning-based heuristic algorithm is developed by embedding a feasibility recovery procedure. Finally, numerical studies demonstrate the superiority of the learning-based heuristic algorithm in both solution optimality and computational efficiency.

关键词： Heuristic algorithms Network topology approximation algorithms Topology Switches Computational efficiency Optimization distribution network reconfiguration optimal transmission switching graph neural network reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

Learner Emotional Value Recognition Technology: An Artificial Neural Network Optimized by the Grey Wolf Algorithm

引用

IEEE ACCESS 2023年 11卷 127689-127696页

作者： Chen, Qi Yang En Univ Law Sch Quanzhou 362014 Peoples R China

In cross-cultural communication, multimedia animation is crucial in defining a nation's image and cultural form. It is a key vehicle for cultural diffusion and a tool for film and television to convey national culture and highlight regional culture. Animation's particular charm, position, and function in cultural dissemination are further highlighted by the special cinematic language used to portray emotions, and it also reflects the medium's unique significance in the rapidly evolving modern society. To better understand the emotions generated by various multimedia animations, in-depth research is needed. To investigate these issues, this article explores the use of the Sigma-pi artificial neural network (SP-ANN) algorithm based on the grey wolf optimization algorithm (GWOA) to identify emotional states. Compared with traditional Sigma n-artificial neural network algorithms, a training process that does not require complex derivative calculations in derivative-based algorithms is performed. Sigma n-networks can benefit from the proposed learning algorithms. This algorithm has high approximation accuracy and is particularly suitable for real-time approximation of nonlinear processes. The test results indicate that the proposed algorithm can work as expected.

关键词： Animation Artificial neural networks Computational modeling Clustering algorithms approximation algorithms Training Mathematical models Cultural aspects Multimedia systems Multimedia animation grey wolf optimization algorithm sigma pi artificial neural network algorithms

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：