检索结果-内蒙古大学图书馆

arXiv 2024年

作者： Park, Sungwoo Kim, Dongjun Alaa, Ahmed M. Department of Electrical Engineering and Computer Sciences UC Berkeley United States Department of Computer Science Stanford United States UCSF United States

In this paper, we introduce a new class of score-based generative models (SGMs) designed to handle high-cardinality data distributions by leveraging concepts from mean-field theory. We present mean-field chaos diffusion models (MF-CDMs), which address the curse of dimensionality inherent in high-cardinality data by utilizing the propagation of chaos property of interacting particles. By treating high-cardinality data as a large stochastic system of interacting particles, we develop a novel score-matching method for infinite-dimensional chaotic particle systems and propose an approximation scheme that employs a subdivision strategy for efficient training. Our theoretical and empirical results demonstrate the scalability and effectiveness of MF-CDMs for managing large high-cardinality data structures, such as 3D point clouds. © 2024, CC BY.

关键词： Stochastic systems

来源：评论

学校读者我要写书评

暂无评论

Qute: Query by Text Search for Time Series Data

Qute: Query by Text Search for Time Series Data

引用

Future Technologies Conference, FTC 2020

作者： Imani, Shima Alaee, Sara Keogh, Eamonn Department of Computer Science and Engineering UC Riverside RiversideCA United States

ISBN: (纸本)9783030630881

Query-based similarity search is a useful exploratory tool that has been used in many areas such as music, economics, and biology to find common patterns and behaviors. Existing query-based search systems allow users to search large time series collections, but these systems are not very robust and they often fail to find similar patterns. In this work, we present Qute (Query by Text) a natural language search framework for finding similar patterns in time series. We show that Qute is expressive while having very small space and time overhead. Qute is a text-based search which leverages information retrieval features such as relevance feedback. Furthermore, Qute subsumes motif and discord/anomaly discovery. We demonstrate the utility of Qute with case studies on both animal behavior and human behavior data. © 2021, Springer Nature Switzerland AG.

关键词： Time series

来源：评论

学校读者我要写书评

暂无评论

Silicon Microring Modulator for High SFDR Analog Links in Monolithic 45nm CMOS

Silicon Microring Modulator for High SFDR Analog Links in Mo...

引用

CLEO: science and Innovations, S and I 2022

作者： Buchbinder, S. Wang, R. Kramnik, D. Van Orden, D. Khilo, A. Fini, J. Sun, C. Wade, M. Stojanović, V. Department of Electrical Engineering and Computer Science UC Berkeley BerkeleyCA94720 United States Ayar Labs EmeryvilleCA94608 United States

We characterize the linearity of a lateral junction microring modulator in a monolithic 45 nm CMOS platform versus modulator bias, optical wavelength, and input power, and achieve peak third-order SFDR of 96.4dB·... 详细信息

ISBN: (纸本)9781557528209

关键词： CMOS integrated circuits

来源：评论

学校读者我要写书评

暂无评论

Towards Data-Driven Policies in Spectrum Management

Towards Data-Driven Policies in Spectrum Management

引用

IEEE International Symposium on New Frontiers in Dynamic Spectrum Access Networks, DySPAN

作者： Karyn Doke Ali Abedi Max Hollingsworth Mariya Zheleva Anant Sahai Dirk Grunwald Keith Gremban Department of Computer Science University at Albany Department of Electrical Engineering and Computer Sciences UC Berkeley Department of Computer Science University of Colorado Boulder Ann and H.J. Smead Aerospace Engineering Sciences University of Colorado Boulder

ISBN: (数字)9798350317640

ISBN: (纸本)9798350317657

This position paper explores the gap between the current state-of-the-art in spectrum management and the objective of data driven spectrum policy. We explore four issues underlying successful data-driven policy: data requirements to support policy decisions; data acquisition and storage; robust, extensible metadata; and tools for analysis and visualization. For each issue, we discuss the state-of-the-art and describe the ultimate objective. We conclude the paper with a call for action to the spectrum community and list a number of efforts that should be undertaken to support true data-driven spectrum policy

关键词： Data acquisition Dynamic spectrum access Data visualization Metadata Radio spectrum management

来源：评论

学校读者我要写书评

暂无评论

A generalized framework for algorithm based team formation

A generalized framework for algorithm based team formation

引用

2021 IEEE International Conference on engineering, Technology and Innovation, ICE/ITMC 2021

作者： Sidhu, Ikhlaq Balakrishnan, Rajarathnam Gopalakrishnan, Sudarshan IEOR SCET UC Berkeley United States IEOR UC Berkeley United States Electrical Engineering and Computer Science UC Berkeley United States

ISBN: (纸本)9781665449632

Teams must be formed for all kinds of projects and purposes. Team formation is a key activity for innovation, entrepreneurship, class projects, and industry initiatives. Our experience with entrepreneurship and innovation has shown that a project's results are highly dependent on the quality of team formation. In real life situations, team building is also time consuming. In this work we are developing a method and approach to form teams using algorithms. To do this, we develop a taxonomy for how teams are formed. We consider the key factors for team formation, identify measures for team effectiveness, as well as synthesize the framework with natural team formation processes. More importantly, we are proposing a generalized framework for algorithm based team formation. We provide sample algorithms with this framework to demonstrate its use in practical situations. The work of this paper also informs future characterization of the models developed here, which includes trade-offs for factors including optimality of team effectiveness, computation time, and policy considerations. © 2021 IEEE.

关键词： Economic and social effects

来源：评论

学校读者我要写书评

暂无评论

LOCAL LIMITS OF SMALL WORLD NETWORKS

arXiv

引用

arXiv 2025年

作者： Alimohammadi, Yeganeh Işik, Senem Saberi, Amin Department of Electrical Engineering and Computer Science UC Berkeley United States Department of Mathematics Stanford University United States Department of Management Science and Engineering Stanford University United States

Small-world networks, known for their high local clustering and short average path lengths, are a fundamental structure in many real-world systems, including social, biological, and technological networks. We apply the theory of local convergence (Benjamini-Schramm convergence) to derive the limiting behavior of the local structures for two of the most commonly studied small-world network models: the Watts-Strogatz model and the Kleinberg model. Establishing local convergence enables us to show that key network measures—such as PageRank, clustering coefficients, and maximum matching size—converge as network size increases, with their limits determined by the graph’s local structure. Additionally, this framework facilitates the estimation of global phenomena, such as information cascades, using local information from small neighborhoods. As an additional outcome of our results, we observe a critical change in the behavior of the limit exactly when the parameter governing long-range connections in the Kleinberg model crosses the threshold where decentralized search remains efficient, offering a new perspective on why decentralized algorithms fail in certain regimes. © 2025, CC BY.

关键词： Small-world networks

来源：评论

学校读者我要写书评

暂无评论

Effectiveness factors for algorithm based team formation with data project case application

Effectiveness factors for algorithm based team formation wit...

引用

2021 IEEE International Conference on engineering, Technology and Innovation, ICE/ITMC 2021

作者： Sidhu, Ikhlaq Gopalakrishnan, Sudarshan Balakrishnan, Rajarathnam IEOR SCET UC Berkeley United States UC Berkeley Electrical Engineering and Computer Science United States IEOR UC Berkeley United States

ISBN: (纸本)9781665449632

Teams must be formed for all kinds of projects and purposes. Team formation is a key activity for innovation, entrepreneurship, class projects, and industry initiatives. In parallel work, we have proposed a generalized framework for Algorithm-based Team Formation. We are interested to apply this framework to the specific task of forming teams for a Data science project course. In this paper, we have focused on the characteristics teams as correlated by the success of the project itself. We find that by examining approximately 30 project teams, there are characteristics which may be used to set feature values for algorithms that can best match student projects together. In particular, teams who work well together with trust and common backgrounds did perform better. Teams with greater domain experience in coding and ML generally also performed better. Surprisingly, diversity in background did not seem to indicate better performance. However, we did observe that individuals who characterize themselves as optimistic and wide comfort zone among others were more strongly present on teams that performed well, speaking to a correlation of individual characteristics/behaviors to team performance. © 2021 IEEE.

关键词： Industries Technological innovation Correlation Conferences Entrepreneurship Data science Approximation algorithms

来源：评论

学校读者我要写书评

暂无评论

Refuting approaches to the log-rank conjecture for XOR functions

arXiv

引用

arXiv 2023年

作者： Hatami, Hamed Hosseini, Kaave Lovett, Shachar Ostuni, Anthony School of Computer Science McGill University Canada Department of Computer Science University of Rochester United States Department of Computer Science and Engineering UC San Diego United States

The log-rank conjecture, a longstanding problem in communication complexity, has persistently eluded resolution for decades. Consequently, some recent efforts have focused on potential approaches for establishing the conjecture in the special case of XOR functions, where the communication matrix is lifted from a boolean function, and the rank of the matrix equals the Fourier sparsity of the function, which is the number of its nonzero Fourier coefficients. In this note, we refute two conjectures. The first has origins in Montanaro and Osborne (arXiv’09) and is considered in Tsang et al. (FOCS’13), and the second one is due to Mande and Sanyal (FSTTCS’20). These conjectures were proposed in order to improve the best-known bound of Lovett (STOC’14) regarding the log-rank conjecture in the special case of XOR functions. Both conjectures speculate that the set of nonzero Fourier coefficients of the boolean function has some strong additive structure. We refute these conjectures by constructing two specific boolean functions tailored to each. © 2023, CC BY.

关键词： Fourier analysis

来源：评论

学校读者我要写书评

暂无评论

Fast exact leverage score sampling from khatri-rao products with applications to tensor decomposition 23

Fast exact leverage score sampling from khatri-rao products ...

引用

Proceedings of the 37th International Conference on Neural Information Processing Systems

作者： Vivek Bharadwaj Osman Asif Malik Riley Murray Laura Grigori Aydin Buluç James Demmel Electrical Engineering and Computer Science Department UC Berkeley and Computational Research Division Lawrence Berkeley National Lab Computational Research Division Lawrence Berkeley National Lab International Computer Science Institute and Computational Research Division Lawrence Berkeley National Lab and Electrical Engineering and Computer Science Department UC Berkeley Institute of Mathematics EPFL & Lab for Simulation and Modelling Paul Scherrer Institute Computational Research Division Lawrence Berkeley National Lab and Electrical Engineering and Computer Science Department UC Berkeley Electrical Engineering and Computer Science Department UC Berkeley

We present a data structure to randomly sample rows from the Khatri-Rao product of several matrices according to the exact distribution of its leverage scores. Our proposed sampler draws each row in time logarithmic in the height of the Khatri-Rao product and quadratic in its column count, with persistent space overhead at most the size of the input matrices. As a result, it tractably draws samples even when the matrices forming the Khatri-Rao product have tens of millions of rows each. When used to sketch the linear least squares problems arising in CANDECOMP / PARAFAC tensor decomposition, our method achieves lower asymptotic complexity per solve than recent state-of-the-art methods. Experiments on billion-scale sparse tensors validate our claims, with our algorithm achieving higher accuracy than competing methods as the decomposition rank grows.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Monitoring AI-Modified Content at Scale: A Case Study on the Impact of ChatGPT on AI Conference Peer Reviews 41

Monitoring AI-Modified Content at Scale: A Case Study on the...

引用

41st International Conference on Machine Learning, ICML 2024

作者： Liang, Weixin Izzo, Zachary Zhang, Yaohui Lepp, Haley Cao, Hancheng Zhao, Xuandong Chen, Lingjiao Ye, Haotian Liu, Sheng Huang, Zhi McFarland, Daniel A. Zou, James Y. Department of Computer Science Stanford University United States Machine Learning Department NEC Labs America United States Department of Electrical Engineering Stanford University United States Graduate School of Education Stanford University United States Department of Management Science and Engineering Stanford University United States Department of Computer Science UC Santa Barbara United States Department of Biomedical Data Science Stanford University United States Department of Sociology Stanford University United States Graduate School of Business Stanford University United States

We present an approach for estimating the fraction of text in a large corpus which is likely to be substantially modified or produced by a large language model (LLM). Our maximum likelihood model leverages expert-written and AI-generated reference texts to accurately and efficiently examine real-world LLM-use at the corpus level. We apply this approach to a case study of scientific peer review in AI conferences that took place after the release of ChatGPT: ICLR 2024, NeurIPS 2023, CoRL 2023 and EMNLP 2023. Our results suggest that between 6.5% and 16.9% of text submitted as peer reviews to these conferences could have been substantially modified by LLMs, i.e. beyond spell-checking or minor writing updates. The circumstances in which generated text occurs offer insight into user behavior: the estimated fraction of LLM-generated text is higher in reviews which report lower confidence, were submitted close to the deadline, and from reviewers who are less likely to respond to author rebuttals. We also observe corpus-level trends in generated text which may be too subtle to detect at the individual level, and discuss the implications of such trends on peer review. We call for future interdisciplinary work to examine how LLM use is changing our information and knowledge practices. Copyright 2024 by the author(s)

关键词： Maximum likelihood

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：