检索结果-内蒙古大学图书馆

11th International Conference on Learning Representations, ICLR 2023

作者： Schuurmans, Jetze Batselier, Kim Kooij, Julian F.P. Netherlands 3mE Faculty TU Delft Netherlands

Tensor decompositions have been successfully applied to compress neural networks. The compression algorithms using tensor decompositions commonly minimize the approximation error on the weights. Recent work assumes the approximation error on the weights is a proxy for the performance of the model to compress multiple layers and fine-tune the compressed model. Surprisingly, little research has systematically evaluated which approximation errors can be used to make choices regarding the layer, tensor decomposition method, and level of compression. To close this gap, we perform an experimental study to test if this assumption holds across different layers and types of decompositions, and what the effect of fine-tuning is. We include the approximation error on the features resulting from a compressed layer in our analysis to test if this provides a better proxy, as it explicitly takes the data into account. We find the approximation error on the weights has a positive correlation with the performance error, before as well as after fine-tuning. Basing the approximation error on the features does not improve the correlation significantly. While scaling the approximation error commonly is used to account for the different sizes of layers, the average correlation across layers is smaller than across all choices (i.e. layers, decompositions, and level of compression) before fine-tuning. When calculating the correlation across the different decompositions, the average rank correlation is larger than across all choices. This means multiple decompositions can be considered for compression and the approximation error can be used to choose between them. © 2023 11th International Conference on Learning Representations, ICLR 2023. All rights reserved.

关键词： approximation algorithms

来源：评论

学校读者我要写书评

暂无评论

Expectation Propagation for Flat-Fading Channels

引用

IEEE WIRELESS COMMUNICATIONS LETTERS 2023年第11期12卷 1846-1850页

作者： Conti, Elisa Piemontese, Amina Colavolpe, Giulio Vannucci, Armando Univ Parma Dept Engn & Architecture I-43124 Parma Italy

This letter addresses the problem of signal detection in flat-fading channels. In this context, receivers based on the expectation propagation framework appear to be very promising although presenting some critical issues. We develop a new algorithm based on this framework where, unlike previous works, convergence is achieved after a single forward-backward pass, without additional inner detector iterations. The proposed message scheduling, together with novel adjustments of the approximating distributions' parameters, allows to obtain significant performance advantages with respect to the state-of-the-art solution. Simulation results show the applicability of this algorithm when sparser pilot configurations have to be adopted and a considerable gain compared to the current available strategies.

关键词： Fading channels Probability density function approximation algorithms Symbols Random variables Convergence Receivers Expectation propagation fading channels parameter estimation single-input single-output (SISO)

来源：评论

学校读者我要写书评

暂无评论

Index Modulation Multiple Access for Unsourced Random Access

引用

IEEE WIRELESS COMMUNICATIONS LETTERS 2023年第5期12卷 794-798页

作者： Su, Yiwei Zheng, Jianping Xidian Univ Sch Telecommun Engn Xian 710071 Shaanxi Peoples R China

In this letter, a novel index modulation multiple access (IMMA) scheme is proposed in the quasi-static MIMO channel. In the proposed scheme, the transmit information bits are divided into two parts, IMMA bits and coded bits. The IMMA bits are employed to select the active compressed sensing codeword transmitted in the first phase, and determine the IMMA pattern. The encoding bits are LDPC encoded, modulated, and transmitted in the second phase through IMMA. In the multiuser detection, the IMMA patterns and active channels are first estimated by approximate message passing (MP) algorithm. Then, the encoding bits are decoded by MP algorithm based on the joint LDPC and IMMA factor graph. Finally, computer simulations are given to demonstrate the performance of the proposed scheme.

关键词： Multiuser detection Indexes Message passing Channel estimation approximation algorithms Phase modulation Symbols Index modulation unsourced random access massive MIMO message passing multiuser detection

来源：评论

学校读者我要写书评

暂无评论

Polynomial-time Combinatorial Algorithm for General Max–Min Fair Allocation

引用

Algorithmica 2024年第2期86卷 485-504页

作者： Ko, Sheng-Yen Chen, Ho-Lin Cheng, Siu-Wing Hon, Wing-Kai Liao, Chung-Shou Department of Industrial Engineering and Engineering Management National Tsing Hua University Hsinchu30013 Taiwan Department of Electrical Engineering National Taiwan University Taipei106 Taiwan Department of Computer Science and Engineering The Hong Kong University of Science and Technology Clear Water Bay Hong Kong Department of Computer Science National Tsing Hua University Hsinchu30013 Taiwan

In the general max–min fair allocation problem, there are m players and n indivisible resources, each player has his/her own utilities for the resources, and the goal is to find an assignment that maximizes the minimum total utility of resources assigned to a player. The problem finds many natural applications such as bandwidth distribution in telecom networks, processor allocation in computational grids, and even public-sector decision making. We introduce an over-estimation strategy to design approximation algorithms for this problem. When all utilities are positive, we obtain an approximation ratio of c1-ϵ, where c is the maximum ratio of the largest utility to the smallest utility of any resource. When some utilities are zero, we obtain an approximation ratio of (1 + 3 c^ + O(δc^ 2)) , where c^ is the maximum ratio of the largest utility to the smallest positive utility of any resource. © 2023, The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature.

关键词： approximation algorithms

来源：评论

学校读者我要写书评

暂无评论

5-approximation for H-Treewidth Essentially as Fast as H-Deletion Parameterized by Solution Size 31

5-Approximation for H-Treewidth Essentially as Fast as H-Del...

引用

31st Annual European Symposium on algorithms, ESA 2023

作者： Jansen, Bart M.P. de Kroon, Jari J.H. Wlodarczyk, Michal Eindhoven University of Technology Netherlands University of Warsaw Poland

ISBN: (纸本)9783959772952

The notion of H-treewidth, where H is a hereditary graph class, was recently introduced as a generalization of the treewidth of an undirected graph. Roughly speaking, a graph of H-treewidth at most k can be decomposed into (arbitrarily large) H-subgraphs which interact only through vertex sets of size O(k) which can be organized in a tree-like fashion. H-treewidth can be used as a hybrid parameterization to develop fixed-parameter tractable algorithms for H-deletion problems, which ask to find a minimum vertex set whose removal from a given graph G turns it into a member of H. The bottleneck in the current parameterized algorithms lies in the computation of suitable tree H-decompositions. We present FPT-approximation algorithms to compute tree H-decompositions for hereditary and union-closed graph classes H. Given a graph of H-treewidth k, we can compute a 5-approximate tree H-decomposition in time f(O(k)) · nO(1) whenever H-deletion parameterized by solution size can be solved in time f(k)·nO(1) for some function f(k) ≥ 2k. The current-best algorithms either achieve an approximation factor of kO(1) or construct optimal decompositions while suffering from non-uniformity with unknown parameter dependence. Using these decompositions, we obtain algorithms solving Odd Cycle Transversal in time 2O(k) · nO(1) parameterized by bipartite-treewidth and Vertex Planarization in time 2O(k log k) · nO(1) parameterized by planar-treewidth, showing that these can be as fast as the solution-size parameterizations and giving the first ETH-tight algorithms for parameterizations by hybrid width measures. © Bart M. P. Jansen, Jari J. H. de Kroon, and Michal Wlodarczyk

关键词： approximation algorithms

来源：评论

学校读者我要写书评

暂无评论

An Implicit Gradient Method for Constrained Bilevel Problems Using Barrier approximation 48

An Implicit Gradient Method for Constrained Bilevel Problems...

引用

48th IEEE International Conference on Acoustics, Speech and Signal Processing, ICASSP 2023

作者： Tsaknakis, Ioannis Khanduri, Prashant Hong, Mingyi University of Minnesota Department of Electrical and Computer Engineering MN United States Wayne State University Department of Computer Science MI United States

ISBN: (纸本)9781728163277

In this work, we propose algorithms for solving a class of Bilevel Optimization (BLO) problems, with applications in areas such as signal processing, networking and machine learning. Specifically, we develop a novel barrier-based gradient approximation algorithm that transforms the constrained BLO problem to a problem with only linear equality constraints in the LL task. For the reformulated problem, we compute the implicit gradient and develop a gradient-based scheme, involving only a single gradient descent step and the (approximate) solution of the linearly constrained strongly convex LL task at each iteration. We establish, under certain assumptions, the non-asymptotic convergence guarantees of the proposed method to stationary points. Finally, we perform a number of experiments that show the potential of the proposed algorithm. © 2023 IEEE.

关键词： approximation algorithms

来源：评论

学校读者我要写书评

暂无评论

Low-Complex Single-Snapshot DoA-Estimation With Higher Degree of Atomic Separation-Freedom in a MU-MIMO System Aided by Prior Information

引用

IEEE SIGNAL PROCESSING LETTERS 2023年 30卷 349-353页

作者： Lau, Ka Ho Ng, Chiu Kit Song, Jian Tsinghua Univ Elect Engn Beijing 10084 Peoples R China Tsinghua Univ Shenzhen Int Grad Sch Shenzhen 518055 Peoples R China

Although Direction-of-Arrival (DoA)-estimation has been well-researched in the recent decades, some specific scenarios with certain requirements are not yet delved in depth. In this letter, we introduce an innovative application for gridless DoA-estimation, called Orthonormal Atomic Norm Minimization (OANM). By applying OANM in case of 1-dimensional DoA-estimation with prior information aided, atomic separation is shown to enjoy a higher degree of freedom by outcoming with a lower DoA-estimation error compared to other methods that are also compressive sensing-based. OANM shows its power by utilizing a single snapshot of the received signal only and estimates all the DoAs of each mobile user in a noisy environment with relatively low error but also low computational complexity when atoms are close in separation.

关键词： Signal processing algorithms Estimation Direction-of-arrival estimation Multiple signal classification Minimization Computational complexity approximation algorithms DoA-estimation compressive sensing norm minimization sparse signal processing orthonormal atomic separation Prony's method single snapshot

来源：评论

学校读者我要写书评

暂无评论

A Mixed Calibration Method of Time-to-Digital Converter for LiDAR Applications

引用

IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS II-EXPRESS BRIEFS 2023年第2期70卷 436-440页

作者： Ma, Jiaji Wang, Xiayu Ma, Rui Hu, Jin Li, Dong Liu, Yang Zhu, Zhangming Xidian Univ Sch Microelect Shaanxi Key Lab Integrated Circuits & Syst Xian 710071 Peoples R China

This brief presents a mixed calibration method of time-to-digital converter (TDC) for Light Detection and Ranging (LiDAR) applications, which is primarily implemented by a feedback loop consisting of VCRO-based TDC, double-digital-to-analog converter (DDAC), encoder, sampling circuit, LSB-first algorithm (LSBFA) block, and analog buffer. Based on the external input frequency control code (FCC), the method automatically adjusts the DDAC output to control TDC. To ensure that the DDAC output meets the accuracy requirements, a voltage coverage strategy is proposed, which is implemented by an external code controlling the R-2R array. Furthermore, after the chip operation is stable, a voltage trimming approach is presented to avoid introducing drastic switching activity, which is implemented using an error comparator and an LSBFA block. The peak-to-peak INL and DNL of TDC are 1.81 LSB and 0.44 LSB, respectively.

关键词： Calibration FCC Codes approximation algorithms Voltage Clocks Laser radar Calibration method TDC LSB-first LiDAR

来源：评论

学校读者我要写书评

暂无评论

Optimizing the Age of Information in RIS-Aided SWIPT Networks

引用

IEEE TRANSACTIONS ON VEHICULAR TECHNOLOGY 2023年第2期72卷 2615-2619页

作者： Lyu, Wanting Xiu, Yue Zhao, Jun Zhang, Zhongpei Univ Elect Sci & Technol China Natl Key Lab Sci & Technol Commun Chengdu 611731 Peoples R China Nanyang Technol Univ Sch Comp Sci & Engn Singapore Singapore Univ Elect Sci & Technol China Shenzhen Inst Adv Study Shenzhen Peoples R China

In this paper, a reconfigurable intelligent surface (RIS)-assisted simultaneous wireless information and power transfer (SWIPT) network is investigated. To quantify the freshness of the data packets at the information receiver, the age of information (AoI) is considered. To minimize the sum AoI of the information users while ensuring that the power transferred to energy harvesting users is greater than the demanded value, we formulate a scheduling scheme, and a joint transmit beamforming and phase shift optimization at the access point (AP) and RIS, respectively. The alternating optimization (AO) algorithm is proposed to handle the coupling between active beamforming and passive RIS phase shifts, and the successive convex approximation (SCA) algorithm is utilized to tackle the non-convexity of the formulated problems. The improvement in terms of AoI provided by the proposed algorithm and the trade-off between the age of information and energy harvesting is quantified by the numerical simulation results.

关键词： Simultaneous wireless information and power transfer Optimization approximation algorithms Array signal processing Wireless sensor networks Receivers Energy harvesting Age of information reconfigurable intelligent surfaces scheduling simultaneous wireless information and power transfer

来源：评论

学校读者我要写书评

暂无评论

Robust Expectation Propagation Detector in Impulsive Noise Channel

引用

IEEE SYSTEMS JOURNAL 2023年第2期17卷 2049-2052页

作者： Dong, Yuanyuan Li, Hua Wang, Xiyuan Dai, Xiaoming Long, Keping Univ Sci & Technol Beijing Sch Comp & Commun Engn Beijing 100083 Peoples R China Res Inst Intelligent Network Zhejiang Lab Hangzhou 311121 Peoples R China Beijing Informat Sci & Technol Univ Sch Informat & Commun Engn Beijing 100101 Peoples R China

In this work, we propose a robust expectation propagation (REP) detector for a system contaminated with impulsive noise. The core idea of the REP is to apply a disperse transformation to the received signal and an associated channel sparsification operation. The proposed disperse transformation distributes the impulsive noise component to multiple physical resource elements. The transformed impulsive noise exhibits greater Gaussianity and better matches the Gaussian assumption of the EP algorithm. The associated channel sparsification can alleviate the adverse effects of deleterious message propagation caused by impulsive noise during the iterative process. Simulation results demonstrate that the proposed REP detector exhibits greater resilience to impulsive noise than the classical EP detector, especially for systems with a high probability of impulsive noise occurrence and/or high-order modulation.

关键词： Detectors Computational complexity Computational modeling Bit error rate approximation algorithms Symbols Sparse matrices Channel sparsification disperse transformation impulsive noise (IN) non-Gaussian robust expectation propagation (REP) detector

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：