检索结果-内蒙古大学图书馆

GEOMETRIC ANALYSIS OF NONCONVEX OPTIMIZATION LANDSCAPES FOR OVERCOMPLETE LEARNING 8

学校读者我要写书评

暂无评论

GEOMETRIC ANALYSIS OF NONCONVEX OPTIMIZATION LANDSCAPES FOR ...

8th International Conference on Learning Representations, ICLR 2020

作者： Qu, Qing Zhai, Yuexiang Li, Xiao Zhang, Yuqian Zhu, Zhihui Center for Data Science New York University United States EECS UC Berkeley United States Eletronic Engineering CUHK Hong Kong Electrical & Computer Engineering Rutgers University United States Electrical & Computer Engineering University of Denver United States

Learning overcomplete representations finds many applications in machine learning and data analytics. In the past decade, despite the empirical success of heuristic methods, theoretical understandings and explanations of these algorithms are still far from satisfactory. In this work, we provide new theoretical insights for several important representation learning problems: learning (i) sparsely used overcomplete dictionaries and (ii) convolutional dictionaries. We formulate these problems as 4-norm optimization problems over the sphere, and study the geometric properties of their nonconvex optimization landscapes. For both problems, we show the nonconvex objectives have benign (global) geometric structures, which enable development of efficient optimization methods finding the target solutions. Finally, our theoretical results are justified by numerical simulations. © 2020 8th International Conference on Learning Representations, ICLR 2020. All rights reserved.

关键词： Heuristic methods

Kernel-based off-policy estimation without overlap: Instance optimality beyond semiparametric efficiency

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Mou, Wenlong Ding, Peng Wainwright, Martin J. Bartlett, Peter L. Department of Electrical Engineering and Computer Sciences United States Department of Statistics UC Berkeley United States Laboratory for Information and Decision Systems EECS and Mathematics Massachusetts Institute of Technology United States Statistics and Data Science Center EECS and Mathematics Massachusetts Institute of Technology United States Google Research United States

We study optimal procedures for estimating a linear functional based on observational data. In many problems of this kind, a widely used assumption is strict overlap, i.e., uniform boundedness of the importance ratio, which measures how well the observational data covers the directions of interest. When it is violated, the classical semi-parametric efficiency bound can easily become infinite, so that the instance-optimal risk depends on the function class used to model the regression function. For any convex and symmetric function class F, we derive a non-asymptotic local minimax bound on the mean-squared error in estimating a broad class of linear functionals. This lower bound refines the classical semi-parametric one, and makes connections to moduli of continuity in functional estimation. When F is a reproducing kernel Hilbert space, we prove that this lower bound can be achieved up to a constant factor by analyzing a computationally simple regression estimator. We apply our general results to various families of examples, thereby uncovering a spectrum of rates that interpolate between the classical theories of semi-parametric efficiency (with √n-consistency) and the slower minimax rates associated with non-parametric function estimation. Copyright © 2023, The Authors. All rights reserved.

关键词： Efficiency

We're Afraid Language Models Aren't Modeling Ambiguity

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Liu, Alisa Wu, Zhaofeng Michael, Julian Suhr, Alane West, Peter Koller, Alexander Swayamdipta, Swabha Smith, Noah A. Choi, Yejin Paul G. Allen School of Computer Science & Engineering University of Washington United States Allen Institute for AI United States University of Southern California United States UC Berkeley United States Saarland University Germany New York University United States Massachusetts Institute of Technology United States

Ambiguity is an intrinsic feature of natural language. Managing ambiguity is a key part of human language understanding, allowing us to anticipate misunderstanding as communicators and revise our interpretations as listeners. As language models are increasingly employed as dialogue interfaces and writing aids, handling ambiguous language is critical to their success. We capture ambiguity in a sentence through its effect on entailment relations with another sentence, and collect AMBIENT, a linguist-annotated benchmark of 1,645 examples with diverse kinds of ambiguity. We design a suite of tests based on AMBIENT, presenting the first evaluation of pretrained LMs to recognize ambiguity and disentangle possible meanings. We find that the task remains extremely challenging, including for GPT-4, whose generated disambiguations are considered correct only 32% of the time in crowdworker evaluation, compared to 90% for disambiguations in our dataset. Finally, to illustrate the value of ambiguity-sensitive tools, we show that a multilabel NLI model can flag political claims in the wild that are misleading due to ambiguity. We encourage the field to rediscover the importance of ambiguity for NLP. © 2023, CC BY.

关键词： Computational linguistics

Physion: Evaluating Physical Prediction from Vision in Humans and Machines 35

学校读者我要写书评

暂无评论

Physion: Evaluating Physical Prediction from Vision in Human...

35th Conference on Neural Information Processing Systems - Track on Datasets and Benchmarks, NeurIPS Datasets and Benchmarks 2021

作者： Bear, Daniel M. Wang, Elias Mrowca, Damian Binder, Felix Tung, Hsiao-Yu Fish Pramod, R.T. Holdaway, Cameron Tao, Sirui Smith, Kevin Sun, Fan-Yun Fei-Fei, Li Kanwisher, Nancy Tenenbaum, Joshua B. Yamins, Daniel L.K. Fan, Judith Department of Psychology StanfordCA94305 United States Department of Electrical Engineering StanfordCA94305 United States Department of Computer Science StanfordCA94305 United States Wu Tsai Neurosciences Institute StanfordCA94305 United States Department of Cognitive Science UC San DiegoCA92093 United States Department of Psychology UC San DiegoCA92093 United States Department of Brain and Cognitive Sciences CBMM MIT CambridgeMA02139 United States

While current vision algorithms excel at many challenging tasks, it is unclear how well they understand the physical dynamics of real-world environments. Here we introduce Physion, a dataset and benchmark for rigorously evaluating the ability to predict how physical scenarios will evolve over time. Our dataset features realistic simulations of a wide range of physical phenomena, including rigid and soft-body collisions, stable multi-object configurations, rolling, sliding, and projectile motion, thus providing a more comprehensive challenge than previous benchmarks. We used Physion to benchmark a suite of models varying in their architecture, learning objective, input-output structure, and training data. In parallel, we obtained precise measurements of human prediction behavior on the same set of scenarios, allowing us to directly evaluate how well any model could approximate human behavior. We found that vision algorithms that learn object-centric representations generally outperform those that do not, yet still fall far short of human performance. On the other hand, graph neural networks with direct access to physical state information both perform substantially better and make predictions that are more similar to those made by humans. These results suggest that extracting physical representations of scenes is the main bottleneck to achieving human-level and human-like physical understanding in vision algorithms. We have publicly released all data and code to facilitate the use of Physion to benchmark additional models in a fully reproducible manner, enabling systematic evaluation of progress towards vision algorithms that understand physical environments as robustly as people do. © 2021 Neural information processing systems foundation. All rights reserved.

关键词： Graph neural networks

IC CLASSIFIER: A CLASSIFIER FOR 3D INDUSTRIAL COMPONENTS BASED ON GEOMETRIC PRIOR USING GNN

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Lin, Zipeng Nie, Zhenguo University of California Berkeley United States Computer Science division UC Berkeley United States Department of Mechanical Engineering Tsinghua University Beijing100084 China State Key Laboratory of Tribology in Advanced Equipment Beijing100084 China Beijing Key Lab of Precision/Ultra-precision Manufacturing Equipments and Control Beijing100084 China

In this paper, we propose an approach to address the problem of classifying 3D industrial components by introducing a novel framework named IC-classifier (Industrial Component classifier). Our framework is designed to focus on the object’s local and global structures, emphasizing the former by incorporating specific local features for embedding the model. By utilizing graphical neural networks and embedding derived from geometric properties, IC-classifier facilitates the exploration of the local structures of the object while using geometric attention for the analysis of global structures. Furthermore, the framework uses point clouds to circumvent the heavy computation workload. The proposed framework’s performance is benchmarked against state-of-the-art models, demonstrating its potential to compete in the field. Copyright © 2023, The Authors. All rights reserved.

关键词： Embeddings

Machine Learning Reveals Composition Dependent Thermal Stability in Halide Perovskites

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Hering, Abigail R. Dubey, Mansha Hosseini, Elahe Srivastava, Meghna An, Yu Correa-Baena, Juan-Pablo Homayoun, Houman Leite, Marina S. Department of Materials Science and Engineering UC Davis Davis95616 United States Department of Electrical and Computer Engineering UC Davis Davis95616 United States School of Materials Science and Engineering Georgia Institute of Technology Atlanta30332 United States Department of Chemistry and Biochemistry Georgia Institute of Technology Atlanta30332 United States

Halide perovskites exhibit unpredictable properties in response to environmental stressors, due to several composition-dependent degradation mechanisms. In this work, we apply data visualization and machine learning (ML) techniques to reveal unexpected correlations between composition, temperature, and material properties while using high throughput, in situ environmental photoluminescence (PL) experiments. Correlation heatmaps show the strong influence of Cs content on film degradation, and dimensionality reduction visualization methods uncover clear composition-based data clusters. An extreme gradient boosting algorithm (XGBoost) effectively forecasts PL features for ten perovskite films with both composition-agnostic (>85% accuracy) and composition-dependent (>75% accuracy) model approaches, while elucidating the relative feature importance of composition (up to 99%). This model validates a previously unseen anti-correlation between Cs content and material thermal stability. Our ML-based framework can be expanded to any perovskite family, significantly reducing the analysis time currently employed to identify stable options for photovoltaics. Copyright © 2025, The Authors. All rights reserved.

关键词： Dimensionality reduction

From tensors to qubits 38

学校读者我要写书评

暂无评论

From tensors to qubits

38th International Symposium on Lattice Field Theory, LATTICE 2021

作者： Meurice, Yannick Bazavov, Alexei Dreher, Patrick Gustafson, Erik Hostetler, Leon Sakai, Ryo Tsai, Shan-Wen Unmuth-Yockey, Judah Zhang, Jin Department of Physics and Astronomy The University of Iowa Iowa CityIA52242 United States Department of Computational Mathematics Science and Engineering Michigan State University East LansingMI48824 United States Department of Physics and Astronomy Michigan State University East LansingMI48824 United States Electrical and Computer Engineering North Carolina State University RaleighNC27695 United States Fermilab BataviaIL60510 United States Department of Physics and Astronomy UC Riverside RiversideCA92521 United States

We discuss recent progress in Tensor Lattice Field Theory and economical, symmetry preserving, truncations suitable for quantum computations/simulations. We focus on spin and gauge models with continuous Abelian symmetries such as the Abelian Higgs model and emphasize noise-robust implementations of Gauss's law. We discuss recent progress concerning the comparison between field digitizations and character expansions, symmetry breaking in tensor language, wave-packet preparation and possible new implementations of Abelian models using Rydberg atoms. © Copyright owned by the author(s) under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License (CC BY-NC-ND 4.0)

关键词： Tensors

Optimizing mixed autonomy traffic flow with decentralized autonomous vehicles and multi-agent RL

学校读者我要写书评

暂无评论

arXiv 2020年

作者： Vinitsky, Eugene Lichtle, Nathan Parvate, Kanaad Bayen, Alexandre M. UC Berkeley Department of Mechanical Engineering UC Berkeley Electrical Engineering and Computer Science UC Berkeley Institute for Transportation Studies Paris-Saclay University ENS Paris-Saclay Department of Computer Science

We study the ability of autonomous vehicles to improve the throughput of a bottleneck using a fully decentralized control scheme in a mixed autonomy setting. We consider the problem of improving the throughput of a scaled model of the San Francisco-Oakland Bay Bridge: a two-stage bottleneck where four lanes reduce to two and then reduce to one. Although there is extensive work examining variants of bottleneck control in a centralized setting, there is less study of the challenging multi-agent setting where the large number of interacting AVs leads to significant optimization difficulties for reinforcement learning methods. We apply multi-agent reinforcement algorithms to this problem and demonstrate that significant improvements in bottleneck throughput, from 20% at a 5% penetration rate to 33% at a 40% penetration rate, can be achieved. We compare our results to a hand-designed feedback controller and demonstrate that our results sharply outperform the feedback controller despite extensive tuning. Additionally, we demonstrate that the RL-based controllers adopt a robust strategy that works across penetration rates whereas the feedback controllers degrade immediately upon penetration rate variation. We investigate the feasibility of both action and observation decentralization and demonstrate that effective strategies are possible using purely local sensing. Finally, we open-source our code at https://***/eugenevinitsky/ decentralized_bottlenecks. Copyright © 2020, The Authors. All rights reserved.

关键词： Feedback control

Standard-Cell Scaling Framework with Guaranteed Pin-Accessibility

学校读者我要写书评

暂无评论

Standard-Cell Scaling Framework with Guaranteed Pin-Accessib...

IEEE International Symposium on Circuits and Systems (ISCAS)

作者： Chung-Kuan Cheng Daeyeal Lee Dongwon Park Computer Science and Engineering UC San Diego La Jolla California Electrical and Computer Engineering UC San Diego La Jolla California

ISBN: (数字)9781728133201

ISBN: (纸本)9781728133218

With the scaling of VLSI technologies, the design-technology co-optimization (DTCO) requires prompt development of standard cell libraries to explore scaling effects of various cell architectures. However, standard cell layout design demands holistic efforts for processing transistor placement and in-cell routing due to the limited routing tracks and complicated design rules. Thus, an automatic design framework of standard cell layout became essential in the advanced scaling. Conventional heuristic/divide-and-conquer approaches lack the optimality of solutions because of the limited solution space. In this paper, we propose a novel standard cell scaling framework that simultaneously finds an optimal solution in placement and routing with the pin-accessibility. To ensure the minimum number of pin-access points, we devise strict Boolean counter-based design constraints. We validate our framework using scaling parameters and cell architectures across sub-7nm technology nodes.

关键词： computer architecture Microprocessors Pins Layout Routing Standards Transistors