检索结果-内蒙古大学图书馆

Boosting black box variational inference 18

学校读者我要写书评

暂无评论

Boosting black box variational inference

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

作者： Francesco Locatello Gideon Dresdner Rajiv Khanna Isabel Valera Gunnar Rätsch Max-Planck Institute for Intelligent Systems Germany and Dept. for Computer Science ETH Zurich Zurich Switzerland. Dept. for Computer Science ETH Zurich Zurich Switzerland The University of Texas at Austin Max-Planck Institute for Intelligent Systems Germany

Approximating a probability density in a tractable manner is a central task in Bayesian statistics. Variational Inference (VI) is a popular technique that achieves tractability by choosing a relatively simple variational approximation. Borrowing ideas from the classic boosting framework, recent approaches attempt to boost VI by replacing the selection of a single density with an iteratively constructed mixture of densities. In order to guarantee convergence, previous works impose stringent assumptions that require significant effort for practitioners. Specifically, they require a custom implementation of the greedy step (called the LMO) for every probabilistic model with respect to an unnatural variational family of truncated distributions. Our work fixes these issues with novel theoretical and algorithmic insights. On the theoretical side, we show that boosting VI satisfies a relaxed smoothness assumption which is sufficient for the convergence of the functional Frank-Wolfe (FW) algorithm. Furthermore, we rephrase the LMO problem and propose to maximize the Residual ELBO (RELBO) which replaces the standard ELBO optimization in VI. These theoretical enhancements allow for black box implementation of the boosting subroutine. Finally, we present a stopping criterion drawn from the duality gap in the classic FW analyses and exhaustive experiments to illustrate the usefulness of our theoretical and algorithmic contributions.

关键词：

The incomplete rosetta stone problem: Identifiability results for multi-view nonlinear ICA

学校读者我要写书评

暂无评论

arXiv 2019年

作者： Gresele, Luigi Rubenstein, Paul K. Mehrjou, Arash Locatello, Francesco Schölkopf, Bernhard Empirical Inference Max Planck Institute for Intelligent Systems Tübingen Germany Max Planck Institute for Biological Cybernetics Tübingen Germany Machine Learning Group University of Cambridge United Kingdom Max Planck ETH Center for Learning Systems Zürich Switzerland BMI Dept. for Computer Science ETH Zürich Switzerland

We consider the problem of recovering a common latent source with independent components from multiple views. This applies to settings in which a variable is measured with multiple experimental modalities, and where the goal is to synthesize the disparate measurements into a single unified representation. We consider the case that the observed views are a nonlinear mixing of component-wise corruptions of the sources. When the views are considered separately, this reduces to nonlinear Independent Component Analysis (ICA) for which it is provably impossible to undo the mixing. We present novel identifiability proofs that this is possible when the multiple views are considered jointly, showing that the mixing can theoretically be undone using function approximators such as deep neural networks. In contrast to known identifiability results for nonlinear ICA, we prove that independent latent sources with arbitrary mixing can be recovered as long as multiple, sufficiently different noisy views are available. Copyright © 2019, The Authors. All rights reserved.

关键词： Mixing

Advances in Intelligent Systems and Computing

学校读者我要写书评

暂无评论

Preface

Advances in intelligent systems and Computing 2021年 1237 AISC卷 v-vi页

作者： Dong, Yucheng Herrera-Viedma, Enrique Matsui, Kenji Omatsu, Shigeru Briones, Alfonso González González, Sara Rodríguez Business School Sichuan University Chengdu China Dept. of System Design Osaka Institute of Technology Osaka Japan GRASIA Research Group Facultad de Informática Universidad Complutense de Madrid Madrid Spain University of Granada Granada Spain Hiroshima University Osaka Japan IoT European Digital Innovation Hub Bioinformatics Intelligent Systems and Educational Technology Research Group Department of Computer Science Faculty of Science University of Salamanca Salamanca Spain

Syncgan: Synchronize the Latent Spaces of Cross-Modal Generative Adversarial Networks

学校读者我要写书评

暂无评论

Syncgan: Synchronize the Latent Spaces of Cross-Modal Genera...

IEEE International Conference on Multimedia and Expo (ICME)

作者： Wen-Cheng Chen Chien-Wen Chen Min-Chun Hu Dept. of Computer Science and Information Engineering National Cheng Kung University Taiwan Graduate Program of Multimedia Systems and Intelligent Computing National Cheng Kung University and Academia Sinica Taiwan

Generative adversarial network (GAN) has achieved impressive success on cross-domain generation, but it faces difficulty in cross-modal generation due to the lack of a common distribution between heterogeneous data. Most existing methods of conditional based cross-modal GANs adopt the strategy of one-directional transfer and have achieved preliminary success on text-to-image transfer. Instead of learning the transfer between different modalities, we aim to learn a synchronous latent space representing the cross-modal common concept. A novel network component named synchronizer is proposed in this work to judge whether the paired data is synchronous/corresponding or not, which can constrain the latent space of generators in the GANs. Our GAN model, named as SyncGAN, can successfully generate synchronous data (e.g., a pair of image and sound) from identical random noise. For transforming data from one modality to another, we recover the latent code by inverting the mappings of a generator and use it to generate data of different modality. In addition, the proposed model can achieve semi-supervised learning, which makes our model more flexible for practical applications.

关键词： Synchronization Gallium nitride Generators Data models Training Generative adversarial networks Training data

Mining within-trial oscillatory brain dynamics to address the variability of optimized spatial filters

学校读者我要写书评

暂无评论

arXiv 2018年

作者： Meinel, Andreas Kolkhorst, Henrich Tangermann, Michael Brain State Decoding Lab Cluster of Excellence BrainLinks-BrainTools Dept. of Computer Science Albert-Ludwigs-University Freiburg Germany Autonomous Intelligent Systems Dept. of Computer Science Albert-Ludwigs-University Freiburg Germany

Data-driven spatial filtering algorithms optimize scores such as the contrast between two conditions to extract oscillatory brain signal components. Most machine learning approaches for filter estimation, however, disregard within-trial temporal dynamics and are extremely sensitive to changes in training data and involved hyperparameters. This leads to highly variable solutions and impedes the selection of a suitable candidate for, e.g., neurotechnological applications. Fostering component introspection, we propose to embrace this variability by condensing the functional signatures of a large set of oscillatory components into homogeneous clusters, each representing specific within-trial envelope dynamics. The proposed method is exemplified by and evaluated on a complex hand force task with a rich within-trial structure. Based on electroencephalography data of 18 healthy subjects, we found that the components’ distinct temporal envelope dynamics are highly subject-specific. On average, we obtained seven clusters per subject, which were strictly confined regarding their underlying frequency bands. As the analysis method is not limited to a specific spatial filtering algorithm, it could be utilized for a wide range of neurotechnological applications, e.g., to select and monitor functionally relevant features for brain-computer interface protocols in stroke rehabilitation. Copyright © 2018, The Authors. All rights reserved.

关键词： Electroencephalography

Competitive training of mixtures of independent deep generative models

学校读者我要写书评

暂无评论

arXiv 2018年

作者： Locatello, Francesco Vincent, Damien Tolstikhin, Ilya Rätsch, Gunnar Gelly, Sylvain Schölkopf, Bernhard Max-Planck Institute for Intelligent Systems Germany BMI Dept. for Computer Science ETH Zurich Switzerland Google AI Brain Team

A common assumption in causal modeling posits that the data is generated by a set of independent mechanisms, and algorithms should aim to recover this structure. Standard unsupervised learning, however, is often concerned with training a single model to capture the overall distribution or aspects thereof. Inspired by clustering approaches, we consider mixtures of implicit generative models that "disentangle" the independent generative mechanisms underlying the data. Relying on an additional set of discriminators, we propose a competitive training procedure in which the models only need to capture the portion of the data distribution from which they can produce realistic samples. As a by-product, each model is simpler and faster to train. We empirically show that our approach splits the training distribution in a sensible way and increases the quality of the generated samples. Copyright © 2018, The Authors. All rights reserved.

关键词： Mixtures

FUTURE-AI: International consensus guideline for trustworthy and deployable artificial intelligence in healthcare

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Lekadir, Karim Feragen, Aasa Fofanah, Abdul Joseph Frangi, Alejandro F. Buyx, Alena Emelie, Anais Lara, Andrea Porras, Antonio R. Chan, An-Wen Navarro, Arcadi Glocker, Ben Botwe, Benard O. Khanal, Bishesh Beger, Brigit Wu, Carol C. Cintas, Celia Langlotz, Curtis P. Rueckert, Daniel Mzurikwao, Deogratias Fotiadis, Dimitrios I. Zhussupov, Doszhan Ferrante, Enzo Meijering, Erik Weicken, Eva González, Fabio A. Asselbergs, Folkert W. Prior, Fred Krestin, Gabriel P. Collins, Gary S. Tegenaw, Geletaw S. Kaissis, Georgios Misuraca, Gianluca Tsakou, Gianna Dwivedi, Girish Kondylakis, Haridimos Jayakody, Harsha Woodruf, Henry C. Mayer, Horst Joachim Aerts, Hugo JWL Walsh, Ian Chouvarda, Ioanna Buvat, Irène Tributsch, Isabell Rekik, Islem Duncan, James Kalpathy-Cramer, Jayashree Zahir, Jihad Park, Jinah Mongan, John Gichoya, Judy W. Schnabel, Julia A. Kushibar, Kaisar Riklund, Katrine Mori, Kensaku Marias, Kostas Amugongo, Lameck M. Fromont, Lauren A. Maier-Hein, Lena Alberich, Leonor Cerdá Rittner, Leticia Phiri, Lighton Marrakchi-Kacem, Linda Donoso-Bach, Lluís Martí-Bonmatí, Luis Cardoso, M. Jorge Bobowicz, Maciej Shabani, Mahsa Tsiknakis, Manolis Zuluaga, Maria A. Bielikova, Maria Fritzsche, Marie-Christine Camacho, Marina Linguraru, Marius George Wenzel, Markus De Bruijne, Marleen Tolsgaard, Martin G. Ghassemi, Marzyeh Ashrafuzzaman, Md Goisauf, Melanie Yaqub, Mohammad Abadía, Mónica Cano Mahmoud, Mukhtar M.E. Elattar, Mustafa Rieke, Nicola Papanikolaou, Nikolaos Lazrak, Noussair Díaz, Oliver Salvado, Olivier Pujol, Oriol Sall, Ousmane Guevara, Pamela Gordebeke, Peter Lambin, Philippe Brown, Pieta Abolmaesumi, Purang Dou, Qi Lu, Qinghua Osuala, Richard Nakasi, Rose Zhou, S. Kevin Napel, Sandy Colantonio, Sara Albarqouni, Shadi Joshi, Smriti Carter, Stacy Klein, Stefan Petersen, Steffen E. Aussó, Susanna Awate, Suyash Raviv, Tammy Riklin Cook, Tessa Mutsvangwa, Tinashe E.M. Rogers, Wendy A. Niessen, Wiro J. Puig-Bosch, Xènia Zeng, Yi Mohammed, Yunusa G. Aquino, Yves Saint James Salahuddin, Zohaib Starmans, Martijn P.A. Department de Matemàtiques i Informàtica Universitat de Barcelona Barcelona Spain Barcelona Spain DTU Compute Technical University of Denmark Kgs Lyngby Denmark Department of Mathematics and Computer Science Faculty of Science and Technology Milton Margai Technical University Freetown Sierra Leone Center for Computational Imaging & Simulation Technologies in Biomedicine Schools of Computing and Medicine University of Leeds Leeds United Kingdom Cardiovascular Science and Electronic Engineering Departments KU Leuven Leuven Belgium Institute of History and Ethics in Medicine Technical University of Munich Munich Germany Faculty of Engineering of Systems Informatics and Sciences of Computing Galileo University Guatemala City Guatemala Department of Biostatistics and Informatics Colorado School of Public Health University of Colorado Anschutz Medical Campus AuroraCO United States Department of Medicine Women’s College Research Institute University of Toronto Toronto Canada Universitat Pompeu Fabra BarcelonaBeta Brain Research Center Barcelona Spain Department of Computing Imperial College London London United Kingdom School of Biomedical & Allied Health Sciences University of Ghana Accra Ghana Department of Midwifery & Radiography School of Health & Psychological Sciences City University of London United Kingdom Kathmandu Nepal European Heart Network Brussels Belgium Department of Thoracic Imaging University of Texas MD Anderson Cancer Center Houston United States IBM Research Africa Nairobi Kenya Departments of Radiology Medicine and Biomedical Data Science Stanford University School of Medicine Stanford United States Institute for AI and Informatics in Medicine Klinikum rechts der Isar Technical University Munich Munich Germany Department of Computing Imperial College London London United Kingdom Muhimbili University of Health and Allied Sciences Dar es Salaam Tanzania United Republic of Ioannina Greece Almaty AI Lab Almaty Kazakhstan

Background: Despite major advances in artificial intelligence (AI) research for healthcare, the deployment and adoption of AI technologies remain limited in clinical practice. In recent years, concerns have been raised about the clinical, technical, ethical and legal risks associated with healthcare AI. To increase adoption in the real world, it is essential that AI technologies are trusted and accepted by patients, clinicians, health organisations and authorities. This paper describes the FUTURE-AI framework as the first international consensus guideline for trustworthy AI in healthcare. Methods: The FUTURE-AI consortium was founded in 2021 and now comprises 117 interdisciplinary experts from 50 countries representing all continents, including AI scientists, clinical researchers, biomedical ethicists, and social scientists. Over a two-year period, the consortium established guiding principles and best practices for trustworthy and deployable AI through an iterative process comprising an in-dept. literature review, a modified Delphi survey, and online consensus meetings. Findings: The FUTURE-AI framework was established based on six guiding principles for trustworthy AI in healthcare, i.e. Fairness, Universality, Traceability, Usability, Robustness and Explainability. Through consensus, a set of 30 best practices were defined, addressing technical, clinical, socio-ethical and legal dimensions of trustworthy AI. The recommendations cover the entire lifecycle of healthcare AI, from design, development and validation to regulation, deployment, and monitoring. Interpretation: FUTURE-AI is a structured, risk-informed framework which provides guidance for constructing healthcare AI tools that will be trusted, deployed and adopted in real-world clinical practice. Researchers are encouraged to take the recommendations into account in proof-of-concept stages to facilitate future translation towards clinical practice of healthcare AI. Copyright © 2023, The Authors. All rights

关键词： Clinical research

Probabilistic deep learning using random sum-product networks

学校读者我要写书评

暂无评论

arXiv 2018年

作者： Peharz, Robert Trapp, Martin Vergari, Antonio Kersting, Kristian Stelzner, Karl Ghahramani, Zoubin Molina, Alejandro Dept. of Engineering University of Cambridge Austrian Research Institute for Artificial Intelligence Max Planck Institute for Intelligent Systems Computer Science Dept TU Darmstadt

The need for consistent treatment of uncertainty has recently triggered increased interest in probabilistic deep learning methods. However, most current approaches have severe limitations when it comes to inference, since many of these models do not even permit to evaluate exact data likelihoods. Sum-product networks (SPNs), on the other hand, are an excellent architecture in that regard, as they allow to efficiently evaluate likelihoods, as well as arbitrary marginalization and conditioning tasks. Nevertheless, SPNs have not been fully explored as serious deep learning models, likely due to their special structural requirements, which complicate learning. In this paper, we make a drastic simplification and use random SPN structures which are trained in a "classical deep learning manner", i.e. employing automatic differentiation, SGD, and GPU support. The resulting models, called RAT-SPNs, yield prediction results comparable to deep neural networks, while still being interpretable as generative model and maintaining well-calibrated uncertainties. This property makes them highly robust under missing input features and enables them to naturally detect outliers and peculiar samples. Copyright © 2018, The Authors. All rights reserved.

关键词： Deep neural networks

On matching pursuit and coordinate descent

学校读者我要写书评

暂无评论

arXiv 2018年

作者： Locatello, Francesco Raj, Anant Karimireddy, Sai Praneeth Rätsch, Gunnar Schölkopf, Bernhard Stich, Sebastian U. Jaggi, Martin Max Planck Institute for Intelligent Systems Tübingen Germany Dept. of Computer Science Eth Zurich Zurich Switzerland Epfl Lausanne Switzerland

Two popular examples of first-order optimization methods over linear spaces are coordinate descent and matching pursuit algorithms, with their randomized variants. While the former targets the optimization by moving along coordinates, the latter considers a generalized notion of directions. Exploiting the connection between the two algorithms, we present a unified analysis of both, providing affine invariant sublinear O(1=t rates on smooth objectives and linear convergence on strongly convex objectives. As a byproduct of our affine invariant analysis of matching pursuit, our rates for steepest coordinate descent are the tightest known. Furthermore, we show the first accelerated convergence rate O(1=t2 for matching pursuit and steepest coordinate descent on convex objectives. Copyright © 2018, The Authors. All rights reserved.

关键词： Machine learning