检索结果-内蒙古大学图书馆

Robust Sea Clutter Suppression by C-ASJU Dictionary learning and ECWT denoised algorithm

学校读者我要写书评

暂无评论

IEEE Sensors Journal 2025年

作者： Zhao, Xinyu Liu, Peng Wang, Bingnan Jin, Yaqiu Fudan University Key Laboratory for Information Science of Electromagnetic Waves Shanghai200433 China Chinese Academy of Sciences National Key Laboratory of Microwave Imaging Technology Aerospace Information Research Institute Beijing100190 China

To address the challenge of detecting weak targets obscured by strong sea clutter and noise in complex maritime environments, this paper presents a robust sea clutter suppression method. It combines a complex-adaptive sequential joint update (C-ASJU) dictionary learning algorithm with an enhanced continuous wavelet threshold (ECWT) denoising technique. First, the proposed dictionary learning (DL) and sparse representation (SR) are utilized to differentiate between sea clutter and target signals. The ASJU dictionary learning algorithm employs rank-1 matrix approximation and adaptive sparse regularization to balance computational efficiency and accuracy. By extending the method to the complex domain, the proposed C-ASJU algorithm can process complex sea clutter signals directly and preserve the phase information by using the Wirtinger calculus, thereby enhancing its applicability in real-time radar systems. To further adapt to diverse oceanic environments, an enhanced continuous wavelet threshold denoising method is integrated to effectively suppress the residual noise and enhance target detection efficiency. Compared with traditional algorithms and other DL algorithms, this method excels in clutter suppression performance, computational efficiency, and significantly boosts weak target detection in sea clutter. Experimental results on the IPIX radar datasets demonstrate the robustness and reliability of the proposed method. © 2001-2012 IEEE.

关键词： learning algorithms

Protaugment: Unsupervised diverse short-texts paraphrasing for intent detection meta-learning

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Dopierre, Thomas Gravier, Christophe Logerais, Wilfried Laboratoire Hubert Curien UMR CNRS 5516 Université Jean Monnet Saint-Étienne France Meetic Paris France

Recent research considers few-shot intent detection as a meta-learning problem: the model is learning to learn from a consecutive set of small tasks named episodes. In this work, we propose PROTAUGMENT, a meta-learning algorithm for short texts classification applied to the intent detection task. PROTAUGMENT is a novel extension of Prototypical Networks (Snell et al., 2017) that limits overfitting on the bias introduced by the few-shots classification objective at each episode. It relies on diverse paraphrasing: a conditional language model is first fine-tuned for paraphrasing, and diversity is later introduced at the decoding stage at each meta-learning episode. The diverse paraphrasing is unsupervised as it is applied to unlabelled data and then fueled to the Prototypical Network training objective as a consistency loss. PROTAUGMENT is the state-of-the-art method for intent detection meta-learning, at no extra labeling efforts and without the need to fine-tune a conditional language model on a given application domain. Copyright © 2021, The Authors. All rights reserved.

关键词： learning algorithms

Rotation excursion algorithm with learning

学校读者我要写书评

暂无评论

arXiv 2025年

作者： He, Sheng-Xue Business School University of Shanghai for Science and Technology Shanghai200093 China

We introduce a novel heuristic algorithm named the Rotation Excursion Algorithm with learning (REAL) designed for general-purpose optimization. REAL draws inspiration from the construction mechanism inherent in CEC optimization suites, integrating three fundamental operations with a natural growth rule to address optimization tasks. The initial operation involves rotating the current feasible solutions within the search space to generate and evaluate new solutions. The excursion operation aims to relocate current feasible solutions closer to historically superior solutions stored in a list known as the "list of visible spots." The third operation involves perturbing solutions generated by the preceding operations within their respective neighborhoods. The rotation operation is geared toward comprehensive and random exploration of the entire search space, while the excursion operation exploits known information to refine current solutions. Perturbation operation functions as a form of neighborhood search to further enhance solution quality. The natural growth rule dynamically adjusts REAL's balance between exploration and exploitation throughout the entire search process. To validate the efficacy of the proposed algorithm, we apply it to address a diverse set of 67 problems, encompassing 29 benchmark optimization problems, 30 test problems from CEC 2014, one from CEC 2022, and seven engineering problems. Numerical experiments demonstrate the superior performance of REAL when compared to various other *** Codes 90 © 2025, CC BY-NC-ND.

关键词： learning algorithms

DIFFERENTIALLY PRIVATE SEQUENTIAL learning

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Liu, Yuxin Rahimian, M. Amin Industrial Engineering University of Pittsburgh United States

In a differentially private sequential learning setting, agents introduce endogenous noise into their actions to maintain privacy. Applying this approach to a standard sequential learning model leads to different outcomes depending on whether the signals are continuous or binary. For continuous signals with a nonzero finite privacy budget, we introduce a novel smoothed randomized response mechanism that adapts the noise level based on the distance to a threshold, unlike the traditional randomized response mechanism, which applies uniform noise. This approach allows agents’ actions to more accurately reflect a combination of their private signals and the observed public history, resulting in an accelerated asymptotic learning speed, reaching Θϵ(log(n)) compared to the Θ(plog(n)) rate in the non-private regime where the privacy budget approaches infinity. Additionally, in the non-private setting, the expected stopping time for achieving the first correct decision and the expected number of incorrect actions both diverge to infinity, meaning that early agents may continue making incorrect decisions for an unreasonably long period. In contrast, under a finite privacy budget ϵ ∈ (0, 1), both of these measures remain finite, suggesting a sharp and surprising contrast between sequential learning outcomes in the private and non-private regimes. learning with continuous signals in the private regime is more efficient because the varying probability of flipping actions between the true and false states under the smooth randomized response amplifies the log-likelihood ratio over time, improving information aggregation between agents. In contrast, when signals are binary, the addition of differential privacy noise consistently hampers the learning performance because agents tend to employ a constant randomized response strategy before an information cascade occurs, reducing the informativeness of their actions and ultimately hindering the overall learning process. Copyright © 2

关键词： learning algorithms

Server-side local gradient averaging and learning rate acceleration for scalable split learning

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Pal, Shraman Uniyal, Mansi Park, Jihong Vepakomma, Praneeth Raskar, Ramesh Bennis, Mehdi Jeon, Moongu Choi, Jinho IIT Kharagpur India Deakin University Australia MIT Media Lab United States GIST Korea Republic of University of Oulu Finland

In recent years, there have been great advances in the field of decentralized learning with private data. Federated learning (FL) and split learning (SL) are two spearheads possessing their pros and cons, and are suited for many user clients and large models, respectively. To enjoy both benefits, hybrid approaches such as SplitFed have emerged of late, yet their fundamentals have still been illusive. In this work, we first identify the fundamental bottlenecks of SL, and thereby propose a scalable SL framework, coined SGLR. The server under SGLR broadcasts a common gradient averaged at the split-layer, emulating FL without any additional communication across clients as opposed to SplitFed. Meanwhile, SGLR splits the learning rate into its server-side and client-side rates, and separately adjusts them to support many clients in parallel. Simulation results corroborate that SGLR achieves higher accuracy than other baseline SL methods including SplitFed, which is even on par with FL consuming higher energy and communication costs. As a secondary result, we observe greater reduction in leakage of sensitive information via mutual information using SLGR over the baselines. Copyright © 2021, The Authors. All rights reserved.

关键词： learning algorithms

AdaptCL: Efficient collaborative learning with dynamic and adaptive pruning

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Zhou, Guangmeng Xu, Ke Li, Qi Liu, Yang Zhao, Yi Department of Computer Science and Technology Tsinghua University Beijing China

In multi-party collaborative learning, the parameter server sends a global model to each data holder for local training and then aggregates committed models globally to achieve privacy protection. However, both the dragger issue of synchronous collaborative learning and the staleness issue of asynchronous collaborative learning make collaborative learning inefficient in real-world heterogeneous environments. We propose a novel and efficient collaborative learning framework named AdaptCL, which generates an adaptive sub-model dynamically from the global base model for each data holder, without any prior information about worker capability. All workers (data holders) achieve approximately identical update time as the fastest worker by equipping them with capability-adapted pruned models. Thus the training process can be dramatically accelerated. Besides, we tailor the efficient pruned rate learning algorithm and pruning approach for AdaptCL. Meanwhile, AdaptCL provides a mechanism for handling the trade-off between accuracy and time overhead and can be combined with other techniques to accelerate training further. Empirical results show that AdaptCL introduces little computing and communication overhead. AdaptCL achieves time savings of more than 41% on average and improves accuracy in a low heterogeneous environment. In a highly heterogeneous environment, AdaptCL achieves a training speedup of 6.2x with a slight loss of accuracy. Copyright © 2021, The Authors. All rights reserved.

关键词： learning algorithms

A novel heuristic algorithm: adaptive and various learning-based algorithm

学校读者我要写书评

暂无评论

arXiv 2025年

作者： He, Sheng-Xue Business School University of Shanghai for Science and Technology Shanghai200093 China

A novel population-based heuristic algorithm called the adaptive and various learning-based algorithm (AVLA) is proposed for solving general optimization problems in this paper. The main idea of AVLA is inspired by the learning behaviors of individuals in a group, e.g. a school class. The algorithm formulates the following learning behaviors: a. Elite members will learn from each other;b. A common member will learn from some elite member and other common members;c. Members with unsatisfied performance will reflect their behavior after performance estimation;d. The whole group will reflect their behavior and try to improve if the performance of the group as a whole has not been improved for a long time. AVLA adopts the success-history based parameter adaptation to lighten the burden of parameter adjustment. To verify the efficiency of the AVLA, we apply it and its no-adaptation version with other eight well-known heuristics to 100 benchmark problems. The comparison clearly shows that AVLA performs as well as SHADE and the non-adaption version of AVLA outperforms others except AVLA and *** Codes 90 © 2025, CC BY-NC-ND.

关键词： learning algorithms

Put CASH on Bandits: A Max K-Armed Problem for Automated Machine learning

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Balef, Amir Rezaei Vernade, Claire Eggensperger, Katharina Department of Computer Science University of Tübingen Germany

The Combined Algorithm Selection and Hyperparameter optimization (CASH) is a challenging resource allocation problem in the field of AutoML. We propose MaxUCB, a max k-armed bandit method to trade off exploring different model classes and conducting hyperparameter optimization. MaxUCB is specifically designed for the light-tailed and bounded reward distributions arising in this setting and, thus, provides an efficient alternative compared to classic max k-armed bandit methods assuming heavy-tailed reward distributions. We theoretically and empirically evaluate our method on four standard AutoML benchmarks, demonstrating superior performance over prior approaches. Copyright © 2025, The Authors. All rights reserved.

关键词： learning algorithms

Unleashing the Power of Continual learning on Non-Centralized Devices: A Survey

学校读者我要写书评

暂无评论

IEEE Communications Surveys and Tutorials 2025年

作者： Li, Yichen Wang, Haozhao Xu, Wenchao Xiao, Tianzhe Liu, Hong Tu, Minzhu Wang, Yuying Yang, Xin Zhang, Rui Yu, Shui Guo, Song Li, Ruixuan Huazhong University of Science and Technology School of Computer Science and Technology Wuhan China The Hong Kong Polytechnic University Department of Computing Hong Kong Beijing China Soochow University School of Computer Science and Technology Suzhou China Southwestern University of Finance and Economics School of Computing and Artificial Intelligence Chengdu China University of Technology Sydney School of Computer Science Sydney Australia The Hong Kong University of Science and Technology Department of Computer Science and Engineering Hong Kong

Non-Centralized Continual learning (NCCL) has become an emerging paradigm for enabling distributed devices such as vehicles and servers to handle streaming data from a joint non-stationary environment. To achieve high reliability and scalability in deploying this paradigm in distributed systems, it is essential to overcome challenges stemming from both spatial and temporal dimensions, manifesting as distribution shifts, catastrophic forgetting, heterogeneity, and privacy issues. This survey focuses on a comprehensive examination of the development of the non-centralized continual learning algorithms and the real-world deployment across distributed devices. We begin with an introduction to the background and fundamentals of non-centralized learning and continual learning. Then, we review existing solutions from three levels to represent how existing techniques alleviate the catastrophic forgetting and distribution shift. Additionally, we delve into the various types of heterogeneity issues, security, and privacy attributes, as well as real-world applications across three prevalent scenarios. Furthermore, we establish a large-scale benchmark to revisit this problem and analyze the performance of the state-of-the-art NCCL approaches. Finally, we discuss the important challenges and future research directions in NCCL. © 2021 IEEE.

关键词： learning algorithms