检索结果-内蒙古大学图书馆

Author Correction: Drug classification with a spectral barcode obtained with a smartphone Raman spectrometer

学校读者我要写书评

暂无评论

Nature communications 2023年第1期14卷 5577页

作者： Un Jeong Kim Suyeon Lee Hyochul Kim Yeongeun Roh Seungju Han Hojung Kim Yeonsang Park Seokin Kim Myung Jin Chung Hyungbin Son Hyuck Choo Metaphotonics TU Samsung Advanced Institute of Technology Suwon Gyeonggi-do 16419 Republic of Korea. Machine Learning TU Samsung Advanced Institute of Technology Suwon Gyeonggi-do 16419 Republic of Korea. Department of Physics Chungnam National University Daejeon 34134 Korea. Institute of Quantum Systems Daejeon 34134 Korea. School of Integrative Engineering Chung-Ang University Seoul 06974 Republic of Korea. Department of Digital Health Samsung Advanced Institute of Health Science Sungkyunkwan University Seoul 06355 Korea. Department of Radiology Samsung Medical Center Sungkyunkwan University Seoul 06355 Korea. Department of Data Convergence and Future Medicine Sungkyunkwan University School of Medicine Suwon Gyeonggi-do 16419 Korea. Medical AI Research Center Research Institute for Future Medicine Samsung Medical Center Seoul 06351 Korea. Metaphotonics TU Samsung Advanced Institute of Technology Suwon Gyeonggi-do 16419 Republic of Korea. hyuck.choo@***.

Bayesian network based label correlation analysis for multi-label classifier chain

学校读者我要写书评

暂无评论

arXiv 2019年

作者： Wang, Ran Ye, Suhe Li, Ke Kwong, Sam College of Mathematics and Statistics Shenzhen University Shenzhen518060 China Shenzhen Key Laboratory of Advanced Machine Learning and Applications Shenzhen University Shenzhen518060 China Department of Computer Science University of Exeter ExeterEX4 4QF United Kingdom Department of Computer Science City University of Hong Kong 83 Tat Chee Avenue Kowloon Hong Kong

Classifier chain (CC) is a multi-label learning approach that constructs a sequence of binary classifiers according to a label order. Each classifier in the sequence is responsible for predicting the relevance of one label. When training the classifier for a label, proceeding labels will be taken as extended features. If the extended features are highly correlated to the label, the performance will be improved, otherwise, the performance will not be influenced or even degraded. How to discover label correlation and determine the label order is critical for CC approach. This paper employs Bayesian network (BN) to model the label correlations and proposes a new BN-based CC method (BNCC). First, conditional entropy is used to describe the dependency relations among labels. Then, a BN is built up by taking nodes as labels and weights of edges as their dependency relations. A new scoring function is proposed to evaluate a BN structure, and a heuristic algorithm is introduced to optimize the BN. At last, by applying topological sorting on the nodes of the optimized BN, the label order for constructing CC model is derived. Experimental comparisons demonstrate the feasibility and effectiveness of the proposed method. Copyright © 2019, The Authors. All rights reserved.

关键词： Bayesian networks

Nonparametric regression with comparisons: Escaping the curse of dimensionality with ordinal information

学校读者我要写书评

暂无评论

arXiv 2018年

作者： Xu, Yichong Muthakana, Hariank Balakrishnan, Sivaraman Dubrawski, Artur Singh, Aarti Machine Learning Department Carnegie Mellon University Pittsburgh United States Department of Statistics and Data Science Carnegie Mellon University Pittsburgh United States Auton Lab Carnegie Mellon University Pittsburgh United States

In supervised learning, we leverage a labeled dataset to design methods for function estimation. In many practical situations, we are able to obtain alternative feedback, possibly at a low cost. A broad goal is to understand the usefulness of, and to design algorithms to exploit, this alternative feedback. We focus on a semi-supervised setting where we obtain additional ordinal (or comparison) information for potentially unlabeled samples. We consider ordinal feedback of varying qualities where we have either a perfect ordering of the samples, a noisy ordering of the samples or noisy pairwise comparisons between the samples. We provide a precise quantification of the usefulness of these types of ordinal feedback in nonparametric regression, showing that in many cases it is possible to accurately estimate an underlying function with a very small labeled set, effectively escaping the curse of dimensionality. We develop an algorithm called Ranking-Regression (R2) and analyze its accuracy as a function of size of the labeled and unlabeled datasets and various noise parameters. We also present lower bounds, that establish fundamental limits for the task and show that R2 is optimal in a variety of settings. Finally, we present experiments that show the efficacy of R2 and investigate its robustness to various sources of noise and model-misspecification. Copyright © 2018, The Authors. All rights reserved.

关键词： Feedback

learning-to-learn stochastic gradient descent with biased regularization

学校读者我要写书评

暂无评论

arXiv 2019年

作者： Denevi, Giulia Ciliberto, Carlo Grazzi, Riccardo Pontil, Massimiliano Computational Statistics and Machine Learning Istituto Italiano di Tecnologia Genoa16163 Italy Department of Mathematics University of Genoa Genoa16146 Italy Department of Electrical and Electronic Engineering Imperial College of London LondonSW7 1AL United Kingdom Department of Computer Science University College London LondonWC1E 6BT United Kingdom

We study the problem of learning-to-learn: inferring a learning algorithm that works well on tasks sampled from an unknown distribution. As class of algorithms we consider Stochastic Gradient Descent on the true risk regularized by the square euclidean distance to a bias vector. We present an average excess risk bound for such a learning algorithm. This result quantifies the potential benefit of using a bias vector with respect to the unbiased case. We then address the problem of estimating the bias from a sequence of tasks. We propose a meta-algorithm which incrementally updates the bias, as new tasks are observed. The low space and time complexity of this approach makes it appealing in practice. We provide guarantees on the learning ability of the meta-algorithm. A key feature of our results is that, when the number of tasks grows and their variance is relatively small, our learning-to-learn approach has a significant advantage over learning each task in isolation by Stochastic Gradient Descent without a bias term. We report on numerical experiments which demonstrate the effectiveness of our approach. Copyright © 2019, The Authors. All rights reserved.

关键词： learning algorithms

Noisy multi-label semi-supervised dimensionality reduction

学校读者我要写书评

暂无评论

arXiv 2019年

作者： Mikalsen, Karl Øyvind Soguero-Ruiz, Cristina Bianchi, Filippo Maria Jenssen, Robert Dept. of Mathematics and Statistics UiT The Arctic University of Norway Tromsø Norway UiT Machine Learning Group Dept. of Signal Theory and Comm. Telematics and Computing Universidad Rey Juan Carlos Fuenlabrada Spain Dept. of Physics and Technology UiT Tromsø Norway Department of Mathematics and Statistics Faculty of Science and Technology UiT – The Arctic University of Norway TromsøN-9037 Norway

Noisy labeled data represent a rich source of information that often are easily accessible and cheap to obtain, but label noise might also have many negative consequences if not accounted for. How to fully utilize noisy labels has been studied extensively within the framework of standard supervised machine learning over a period of several decades. However, very little research has been conducted on solving the challenge posed by noisy labels in non-standard settings. This includes situations where only a fraction of the samples are labeled (semi-supervised) and each high-dimensional sample is associated with multiple labels. In this work, we present a novel semi-supervised and multi-label dimensionality reduction method that effectively utilizes information from both noisy multi-labels and unlabeled data. With the proposed Noisy multi-label semi-supervised dimensionality reduction (NMLSDR) method, the noisy multi-labels are denoised and unlabeled data are labeled simultaneously via a specially designed label propagation algorithm. NMLSDR then learns a projection matrix for reducing the dimensionality by maximizing the dependence between the enlarged and denoised multi-label space and the features in the projected space. Extensive experiments on synthetic data, benchmark datasets, as well as a real-world case study, demonstrate the effectiveness of the proposed algorithm and show that it outperforms state-of-the-art multi-label feature extraction algorithms. Copyright © 2019, The Authors. All rights reserved.

关键词： Supervised learning

Low rank approximation of binary matrices: Column subset selection and generalizations 43

学校读者我要写书评

暂无评论

Low rank approximation of binary matrices: Column subset sel...

43rd International Symposium on Mathematical Foundations of Computer science, MFCS 2018

作者： Dan, Chen Hansen, Kristoffer Arnsfelt Jiang, He Wang, Liwei Zhou, Yuchen Carnegie Mellon University PittsburghPA United States Department of Computer Science Aarhus University Aarhus Denmark University of Southern California Los AngelesCA United States Key Laboratory of Machine Perception MOE School of EECS Peking University China Center for Data Science Peking University Beijing Institute of Big Data Research Beijing China Department of Statistics University of Wisconsin-Madison MadisonWI United States

ISBN: (纸本)9783959770866

Low rank approximation of matrices is an important tool in machine learning. Given a data matrix, low rank approximation helps to find factors, patterns, and provides concise representations for the data. Research on low rank approximation usually focuses on real matrices. However, in many applications data are binary (categorical) rather than continuous. This leads to the problem of low rank approximation of binary matrices. Here we are given a d × n binary matrix A and a small integer k k2 + 1 +2(2kk−1) and this is asymptotically tight. For the Boolean model, it turns out that CSS is no longer sufficient to obtain a bound. We then develop a Generalized CSS (GCSS) procedure in which the columns of U are generated from Boolean formulas operating bitwise on selected columns of the data matrix. We show that the approximation ratio achieved by GCSS is bounded by 2k−1 + 1, and argue that an exponential dependency on k is seems inherent. © Chen Dan, Kristoffer Arnsfelt Hansen, He Jiang, Liwei Wang, and Yuchen Zhou.

关键词： Approximation algorithms

Algorithmic regularization in learning deep homogeneous models: layers are automatically balanced 18

学校读者我要写书评

暂无评论

Algorithmic regularization in learning deep homogeneous mode...

Proceedings of the 32nd International Conference on Neural Information Processing Systems

作者： Simon S. Du Wei Hu Jason D. Lee Machine Learning Department School of Computer Science Carnegie Mellon University Computer Science Department Princeton University Department of Data Sciences and Operations Marshall School of Business University of Southern California

We study the implicit regularization imposed by gradient descent for learning multi-layer homogeneous functions including feed-forward fully connected and convolutional deep neural networks with linear, ReLU or Leaky ReLU activation. We rigorously prove that gradient flow (i.e. gradient descent with infinitesimal step size) effectively enforces the differences between squared norms across different layers to remain invariant without any explicit regularization. This result implies that if the weights are initially small, gradient flow automatically balances the magnitudes of all layers. Using a discretization argument, we analyze gradient descent with positive step size for the non-convex low-rank asymmetric matrix factorization problem without any regularization. Inspired by our findings for gradient flow, we prove that gradient descent with step sizes ηt = O(t-(1/2 + δ)) (0 < δ ≤ 1/2) automatically balances two low-rank factors and converges to a bounded global optimum. Furthermore, for rank-1 asymmetric matrix factorization we give a finer analysis showing gradient descent with constant step size converges to the global minimum at a globally linear rate. We believe that the idea of examining the invariance imposed by first order algorithms in learning homogeneous models could serve as a fundamental building block for studying optimization for learning deep models.

关键词：

Genome-wide association study of long COVID

学校读者我要写书评

暂无评论

Nature Genetics 2025年 1-16页

作者： Lammi, Vilma Nakanishi, Tomoko Jones, Samuel E. Andrews, Shea J. Karjalainen, Juha Cortés, Beatriz O’Brien, Heath E. Ochoa-Guzman, Ana Fulton-Howard, Brian E. Broberg, Martin Haapaniemi, Hele H. Kanai, Masahiro Pirinen, Matti Schmidt, Axel Mitchell, Ruth E. Mousas, Abdou Mangino, Massimo Huerta-Chagoya, Alicia Sinnott-Armstrong, Nasa Cirulli, Elizabeth T. Vaudel, Marc Kwong, Alex S. F. Maiti, Amit K. Marttila, Minttu M. Posner, Daniel C. Rodriguez, Alexis A. Batini, Chiara Minnai, Francesca Dearman, Anna R. Warmerdam, C. A. Robert Sequeros, Celia B. Winkler, Thomas W. Jordan, Daniel M. Rešcenko, Raimonds Miano, Lorenzo Lane, Jacqueline M. Chung, Ryan K. Guillen-Guio, Beatriz Leavy, Olivia C. Carvajal-Silva, Laura Aguilar-Valdés, Kevin Frangione, Erika Guare, Lindsay Vergasova, Ekaterina Marouli, Eirini Striano, Pasquale Zainulabid, Ummu Afeera Kumar, Ashutosh Ahmad, Hajar Fauzan Edahiro, Ryuya Azekawa, Shuhei Maj, Carlo Pensel, Max C. Miller, Abigail Nöthen, Markus M. Behzad, Pari Schultz, Sonja Heggemann, Julia Balla, Daniella Morrison, David R. Mooser, Vincent Perley, Danielle Ragoussis, Jiannis Rousseau, Simon Shi, Fangyi Lathrop, G. Mark Yanishevsky, Solomia Laurent, Laetitia Auld, Daniel Bertrand, Mylene Boisclair, Ariane Bourque, Guillaume Bujold, David Durand, Madeleine Adra, Darin St-Cyr, Janick Niemi, Mari E. K. Iyengar, Sudha K. Cai, Tianxi Cho, Kelly Sarvadhavabhatla, Sannidhi Donaire, Maria Sophia Valenzuela-Jorquera, Héctor Lamoza, Eduardo Arévalo, Tamara V. Retamales-Ortega, Rocío Sanhueza, Sergio Selman, Carolina S. Silva, Andrea X. Alarcon, Teresa A. Martínez, Matías F. Quiroga, Romina Zuñiga-Pacheco, Paula Zapata-Contreras, Daniela Yáñez, Cristian E. Quiñones, Luis A. Tobar-Calfucoy, Eduardo A. Monardes-Ramírez, Virginia A. Signore, Iskra A. Guajardo, Lissette G. Figueroa, Alvaro Oróstica, Karen Y. Muñoz, Christian A. Fuentes-Guajardo, Macarena Bocchieri, Pamela Cabrera, Camilo Echeverria, Cesar A. Cerpa, Leslie C. Donoso, Gerardo Gutiérrez-Richards, Scarlett Nova-Lamperti, Estefania Saez Hidalgo Institute for Molecular Medicine Finland (FIMM) Helsinki Institute of Life Science (HiLIFE) University of Helsinki Helsinki Finland Department of Human Genetics McGill University Montreal QC Canada Centre for Clinical Epidemiology Department of Medicine Lady Davis Institute Jewish General Hospital McGill University Montreal QC Canada Kyoto-McGill International Collaborative Program in Genomic Medicine Graduate School of Medicine Kyoto University Kyoto Japan Department of Genome Informatics Graduate School of Medicine the University of Tokyo Tokyo Japan Research Fellow Japan Society for the Promotion of Science Tokyo Japan Department of Psychiatry and Behavioral Sciences University of California San Francisco San Francisco CA United States Program in Medical and Population Genetics Broad Institute of Harvard and MIT Cambridge MA United States Stanley Center for Psychiatric Research Broad Institute of Harvard and MIT Cambridge MA United States Analytic and Translational Genetics Unit Massachusetts General Hospital Boston MA United States Genomes for Life-GCAT Lab CORE Program Germans Trias i Pujol Research Institute (IGTP) Badalona Spain Grup de REcerca en Impacte de les Malalties Cròniques i les seves Trajectòries (GRIMTra) Barcelona Spain Sano Genetics Limited London United Kingdom Unidad de Biología Molecular y Medicina Genómica Instituto Nacional de Ciencias Médicas y Nutrición Salvador Zubirán Mexico City Mexico Genetics and Genomic Sciences Icahn School of Medicine at Mount Sinai New York City NY United States Broad Institute Cambridge MA United States Analytical and Translational Genetics Unit Massachusetts General Hospital Harvard Medical School Boston MA United States Department of Mathematics and Statistics University of Helsinki Helsinki Finland Department of Public Health University of Helsinki Helsinki Finland Institute of Human Genetics University of Bonn School of Medicine and University Hospital Bonn Bonn Germany MRC Integrat

Infections can lead to persistent symptoms and diseases such as shingles after varicella zoster or rheumatic fever after streptococcal infections. Similarly, severe acute respiratory syndrome coronavirus 2 (SARS‑CoV‑2) infection can result in long coronavirus disease (COVID), typically manifesting as fatigue, pulmonary symptoms and cognitive dysfunction. The biological mechanisms behind long COVID remain unclear. We performed a genome-wide association study for long COVID including up to 6,450 long COVID cases and 1,093,995 population controls from 24 studies across 16 countries. We discovered an association of FOXP4 with long COVID, independent of its previously identified association with severe COVID-19. The signal was replicated in 9,500 long COVID cases and 798,835 population controls. Given the transcription factor FOXP4’s role in lung physiology and pathology, our findings highlight the importance of lung function in the pathophysiology of long COVID. © The Author(s) 2025.

关键词：

Gradient descent finds global minima of deep neural networks

学校读者我要写书评

暂无评论

arXiv 2018年

作者： Du, Simon S. Lee, Jason D. Li, Haochuan Wang, Liwei Zhai, Xiyu Machine Learning Department Carnegie Mellon University Data Science and Operations Department University of Southern California School of Physics Peking University Center for Data Science Peking University Beijing Institute of Big Data Research Key Laboratory of Machine Perception Moe School of Eecs Peking University Department of Eecs Massachusetts Institute of Technology

Gradient descent finds a global minimum in training deep neural networks despite the objective function being non-convex. The current paper proves gradient descent achieves zero training loss in polynomial time for a deep over-parameterized neural network with residual connections (ResNet). Our analysis relies on the particular structure of the Gram matrix induced by the neural network architecture. This structure allows us to show the Gram matrix is stable throughout the training process and this stability implies the global optimality of the gradient descent algorithm. We further extend our analysis to deep residual convolutional neural networks and obtain a similar convergence result. Copyright © 2018, The Authors. All rights reserved.

关键词： Gradient methods