Probabilistic Automated Numerical Analysis in Machine learning and Artificial intelligence

Calibrated Adaptive Probabilistic ODE Solvers

Autores: Bosch, Nathanael; Hennig, Philipp; Tronarp, Filip
Publicado en: International Conference on Artificial Intelligence and Statistics, Edición 130, 2021, Página(s) 3466-3474
Editor: PMLR
DOI: 10.48550/arxiv.2012.08202

Counterfactual mean embeddings

Autores: Krikamol Muandet, Motonobu Kanagawa, Sorawit Saengkyongam, Sanparith Marukatat
Publicado en: The Journal of Machine Learning Research, Edición 22 (1), 2021, Página(s) 7322-7392
Editor: The Journal of Machine Learning Research

BackPACK: Packing more into backprop

Autores: Dangel, Felix; Kunstner, Frederik; Hennig, Philipp
Publicado en: ICLR, Edición 8, 2020
Editor: ICLR

High-Dimensional Gaussian Process Inference with Derivatives

Autores: de Roos, Filip; Gessner, Alexandra; Hennig, Philipp
Publicado en: International Conference on Machine Learning, Edición 139, 2021, Página(s) 2535-2545
Editor: PMLR

Probabilistic ODE solutions in millions of dimensions

Autores: Nicholas Krämer, Nathanael Bosch, Jonathan Schmidt, Philipp Hennig
Publicado en: International Conference on Machine Learning, Edición 162, 2022, Página(s) 11634-11649
Editor: PMLR

Model-based Kernel Sum Rule: Kernel Bayesian Inference with Probabilistic Models

Autores: Nishiyama, Yu; Kanagawa, Motonobu; Gretton, Arthur; Fukumizu, Kenji
Publicado en: Machine Learning, Edición 109, 2020, Página(s) 939–972
Editor: Springer Link

Probabilistic Numerical Method of Lines for Time-Dependent Partial Differential Equations

Autores: Nicholas Krämer, Jonathan Schmidt, Philipp Hennig
Publicado en: Proceedings of The 25th International Conference on Artificial Intelligence and Statistics, Edición 151, 2022, Página(s) 625-639
Editor: PMLR

An Infinite-Feature Extension for Bayesian ReLU Nets That Fixes Their Asymptotic Overconfidence

Autores: Kristiadi, Agustinus; Hein, Matthias; Hennig, Philipp
Publicado en: Advances in Neural Information Processing Systems, Edición 34, 2021, Página(s) 18789-18800
Editor: Curran Associates Inc
DOI: 10.48550/arxiv.2010.02709

Integrals over Gaussians under Linear Domain Constraints

Autores: Gessner, Alexandra; Kanjilal, Oindrila; Hennig, Philipp
Publicado en: Proceedings of Machine Learning Research, 2020, Página(s) 2764-2774
Editor: PMLR

Convergence Guarantees for Adaptive Bayesian Quadrature Methods

Autores: Kanagawa, Motonobu; Hennig, Philipp
Publicado en: Advances in Neural Information Processing Systems (NeurIPS 2019), Edición 32, 2019
Editor: Advances in Neural Information Processing Systems 32 (NeurIPS 2019)
DOI: 10.48550/arxiv.1905.10271

Descending through a Crowded Valley - Benchmarking Deep Learning Optimizers

Autores: Schmidt, Robin M.; Schneider, Frank; Hennig, Philipp
Publicado en: International Conference on Machine Learning, Edición 139, 2021, Página(s) 9367-9376
Editor: PMLR
DOI: 10.48550/arxiv.2007.01547

Model Selection for Simulator-based Statistical Models: A Kernel Approach

Autores: Kajihara, Takafumi; Kanagawa, Motonobu; Nakaguchi, Yuuki; Khandelwal, Kanishka; Fukumiziu, Kenji
Publicado en: Machine Learning, Edición 109, 2020, Página(s) 939–972
Editor: Springer
DOI: 10.48550/arxiv.1902.02517

Fenrir: Physics-Enhanced Regression for Initial Value Problems

Autores: Filip Tronarp, Nathanael Bosch, Philipp Hennig
Publicado en: International Conference on Machine Learning, Edición 162, 2022, Página(s) 21776--21794
Editor: PMLR

DeepOBS: A Deep Learning Optimizer Benchmark Suite

Autores: Schneider, Frank; Balles, Lukas; Hennig, Philipp
Publicado en: International Conference on Learning Representations, 2019
Editor: International Conference on Learning Representations
DOI: 10.48550/arxiv.1903.05499

The Geometry of Sign Gradient Descent

Autores: Balles, Lukas; Pedregosa, Fabian; Le Roux, Nicolas
Publicado en: ICLR 2020, 2020
Editor: ICLR

Fast Predictive Uncertainty for Classification with Bayesian Deep Networks

Autores: Hobbhahn, Marius; Kristiadi, Agustinus; Hennig, Philipp
Publicado en: Uncertainty in Artificial Intelligence, Edición 180, 2022, Página(s) 822-832
Editor: PMLR
DOI: 10.48550/arxiv.2003.01227

Preconditioning for scalable Gaussian process hyperparameter optimization

Autores: Jonathan Wenger, Geoff Pleiss, Philipp Hennig, John Cunningham, Jacob Gardner
Publicado en: International Conference on Machine Learning, Edición 162, 2022, Página(s) 23751-23780
Editor: PMLR

Learnable Uncertainty under Laplace Approximations

Autores: Kristiadi, Agustinus; Hein, Matthias; Hennig, Philipp
Publicado en: Proceedings of the Thirty-Seventh Conference on Uncertainty in Artificial Intelligence, Edición 161, 2021, Página(s) 344-353
Editor: PMLR
DOI: 10.48550/arxiv.2010.02720

Laplace Redux -- Effortless Bayesian Deep Learning

Autores: Daxberger, Erik; Kristiadi, Agustinus; Immer, Alexander; Eschenhagen, Runa; Bauer, Matthias; Hennig, Philipp
Publicado en: Advances in Neural Information Processing Systems (NeurIPS), Edición 34, 2021, Página(s) 20089-20103
Editor: Advances in Neural Information Processing Systems (NeurIPS)
DOI: 10.48550/arxiv.2106.14806

Probabilistic solvers enable a straight-forward exploration of numerical uncertainty in neuroscience models

Autores: Jonathan Oesterle, Nicholas Krämer, Philipp Hennig, Philipp Berens
Publicado en: Journal of Computational Neuroscience, Edición 50 (4), 2022, Página(s) 485-503
Editor: Springer US
DOI: 10.1007/s10827-022-00827-7

Being Bayesian, Even Just a Bit, Fixes Overconfidence in ReLU Networks

Autores: Kristiadi, Agustinus; Hein, Matthias; Hennig, Philipp
Publicado en: ICML, Edición 4, 2019
Editor: ICML

Probabilistic Linear Solvers for Machine Learning

Autores: Wenger, Jonathan; Hennig, Philipp
Publicado en: Advances in Neural Information Processing Systems, Edición 33, 2020, Página(s) 6731 - 6742
Editor: Curran Associate Inc.
DOI: 10.48550/arxiv.2010.09691

Limitations of the empirical Fisher approximation for natural gradient descent

Autores: Kunstner, Frederik; Hennig, Philipp; Balles, Lukas
Publicado en: Advances in Neural Information Processing Systems 32, Edición 32, 2019, Página(s) {4158--4169
Editor: Curran Associates, Inc.

Pick-and-mix information operators for probabilistic ODE solvers

Autores: Nathanael Bosch, Filip Tronarp, Philipp Hennig
Publicado en: International Conference on Artificial Intelligence and Statistics, Edición 151, 2022, Página(s) 10015-10027
Editor: PMLR

A Fourier State Space Model for Bayesian ODE Filters

Autores: Kersting, Hans; Mahsereci, Maren
Publicado en: Workshop on Invertible Neural Networks, Normalizing Flows, and Explicit Likelihood Models, ICML, 2020
Editor: ICML

Cockpit: A Practical Debugging Tool for the Training of Deep Neural Networks

Autores: Schneider, Frank; Dangel, Felix; Hennig, Philipp
Publicado en: Advances in Neural Information Processing Systems, Edición 34, 2021, Página(s) 20825-20837
Editor: Curran Associates Inc.
DOI: 10.48550/arxiv.2102.06604

Kernel Recursive ABC: Point Estimation with Intractable Likelihood

Autores: Takafumi Kajihara, Motonobu Kanagawa, Keisuke Yamazaki, and Kenji Fukumizu
Publicado en: Proceedings of the International Conference on Machine Learning, Edición 35, 2018, Página(s) 2400-2409
Editor: PMLR (Proceedings of Machine Learning Research

Convergence Guarantees for Adaptive Bayesian Quadrature Methods

Autores: Motonobu Kanagawa, Philipp Hennig
Publicado en: Advances in Neural Information Processing Systems (NeurIPS), Edición 32, 2019, Página(s) 6234--6245
Editor: Curran Associates, Inc.

Dissecting Adam: The Sign, Magnitude and Variance of Stochastic Gradients

Autores: Lukas Balles, Philipp Hennig
Publicado en: Proceedings of the 35th International Conference on Machine Learning (ICML), Edición 35, 2018, Página(s) 404--413
Editor: PMLR

DeepOBS: A Deep Learning Optimizer Benchmark Suite

Autores: Schneider, Frank; Balles, Lukas; Hennig, Philipp
Publicado en: International Conference on Learning Representations (ICLR), Edición 7, 2019
Editor: ICLR

Limitations of the empirical Fisher approximation for natural gradient descent

Autores: Frederik Kunstner, Philipp Hennig, Lukas Balles
Publicado en: Advances in Neural Information Processing Systems (NeurIPS), Edición 32, 2019, Página(s) 4158--4169
Editor: Curran Associates, Inc.

Active Probabilistic Inference on Matrices for Pre-Conditioning in Stochastic Optimization

Autores: Filip de Roos, Philipp Hennig
Publicado en: Proceedings of the 22nd International Conference on Artificial Intelligence and Statistics (AISTATS), Edición 22, 2019, Página(s) 1448--1457
Editor: PMLR

Fast and Robust Shortest Paths on Manifolds Learned from Data

Autores: Georgios Arvanitidis, Soren Hauberg, Philipp Hennig, Michael Schober
Publicado en: Proceedings of the 22nd International Conference on Artificial Intelligence and Statistics (AISTATS), Edición 22, 2019, Página(s) 1506--1515
Editor: JMLR

Active Multi-Information Source Bayesian Quadrature

Autores: Alexandra Gessner, Javier Gonzalez, Maren Mahsereci
Publicado en: Conference on Uncertainty in Artificial Intelligence (UAI), Edición 35, 2019
Editor: UAI

Convergence rates of Gaussian ODE filters

Autores: Hans Kersting; Timothy Sullivan; Philipp Hennig
Publicado en: Statistics and computing, Edición 30 (6), 2020, Página(s) 1791-1816
Editor: Springer US
DOI: 10.1007/s11222-020-09972-4

Being a Bit Frequentist Improves Bayesian Neural Networks

Autores: Kristiadi, Agustinus; Hein, Matthias; Hennig, Philipp
Publicado en: International Conference on Artificial Intelligence and Statistics, Edición 151, 2022, Página(s) 529-545
Editor: PMLR
DOI: 10.48550/arxiv.2106.10065

Linear-Time Probabilistic Solutions of Boundary Value Problems

Autores: Krämer, Nicholas; Hennig, Philipp
Publicado en: Advances in Neural Information Processing Systems 34 (NeurIPS 2021), Edición 34, 2021, Página(s) 11160-11171
Editor: Advances in Neural Information Processing Systems 34 (NeurIPS 2021)

Probabilistic DAG Search

Autores: Grosse, Julia; Zhang, Cheng; Hennig, Philipp
Publicado en: Uncertainty in Artificial Intelligence, Edición 161, 2021, Página(s) 1424-1433
Editor: PMLR
DOI: 10.48550/arxiv.2106.08717

A Probabilistic State Space Model for Joint Inference from Differential Equations and Data

Autores: Schmidt, Jonathan; Krämer, Nicholas; Hennig, Philipp
Publicado en: Advances in Neural Information Processing Systems 34 (NeurIPS 2021), Edición 34, 2021, Página(s) 12374-12385
Editor: Advances in Neural Information Processing Systems 34 (NeurIPS 2021)
DOI: 10.48550/arxiv.2103.10153

ProbNum: Probabilistic Numerics in Python

Autores: Jonathan Wenger, Nicholas Krämer, Marvin Pförtner, Jonathan Schmidt, Nathanael Bosch, Nina Effenberger, Johannes Zenn, Alexandra Gessner, Toni Karvonen, François-Xavier Briol, Maren Mahsereci, Philipp Hennig
Publicado en: 2021
Editor: arXiv
DOI: 10.48550/arxiv.2112.02100

Bayesian Quadrature on Riemannian Data Manifolds

Autores: Fröhlich, Christian; Gessner, Alexandra; Hennig, Philipp; Schölkopf, Bernhard; Arvanitidis, Georgios
Publicado en: International Conference on Machine Learning, Edición 139, 2021, Página(s) 3459-3468
Editor: PMLR
DOI: 10.48550/arxiv.2102.06645

Integrals over Gaussians under Linear Domain Constraints

Autores: Gessner, Alexandra; Kanjilal, Oindrila; Hennig, Philipp
Publicado en: International Conference on Artificial Intelligence and Statistics. Proceedings of Machine Learning Research (PMLR), Edición 1, 2020
Editor: MLR Press
DOI: 10.48550/arxiv.1910.09328

Modular Block-diagonal Curvature Approximations for Feedforward Architectures

Autores: Dangel, Felix; Harmeling, Stefan; Hennig, Philipp
Publicado en: Proceedings of the Twenty Third International Conference on Artificial Intelligence and Statistics, Edición 2, 2020, Página(s) 799-808
Editor: PMLR

Differentiable Likelihoods for Fast Inversion of 'Likelihood-Free' Dynamical Systems

Autores: Kersting, Hans; Krämer, Nicholas; Schiegg, Martin; Daniel, Christian; Tiemann, Michael; Hennig, Philipp
Publicado en: International Conference on Machine Learning (ICML), Edición 11, 2020
Editor: ICML

Resnet after all: Neural odes and their numerical solution

Autores: Katharina Ott, Prateek Katiyar, Philipp Hennig, Michael Tiemann
Publicado en: International Conference on Learning Representations, 2021
Editor: International Conference on Learning Representations

Conjugate Gradients for Kernel Machines

Autores: Bartels, Simon; Hennig, Philipp
Publicado en: Journal of Machine Learning Research, 2020, Página(s) 1-42
Editor: Journal of Machine Learning Research

Bayesian ODE solvers: the maximum a posteriori estimate

Autores: Filip Tronarp; Simo Särkkä; Philipp Hennig
Publicado en: Statistics and Computing, 31(3), Edición 31, 2021, ISSN 0960-3174
Editor: Kluwer Academic Publishers
DOI: 10.1007/s11222-021-09993-7

Convergence Analysis of Deterministic Kernel-Based Quadrature Rules in Misspecified Settings

Autores: Motonobu Kanagawa, Bharath K. Sriperumbudur, Kenji Fukumizu
Publicado en: Foundations of Computational Mathematics, 2019, Página(s) 1-40, ISSN 1615-3375
Editor: Springer Verlag
DOI: 10.1007/s10208-018-09407-7

On the positivity and magnitudes of Bayesian quadrature weights

Autores: Toni Karvonen, Motonobu Kanagawa, Simo Särkkä
Publicado en: Statistics and Computing, Edición 29/6, 2019, Página(s) 1317-1333, ISSN 0960-3174
Editor: Kluwer Academic Publishers
DOI: 10.1007/s11222-019-09901-0

Probabilistic solutions to ordinary differential equations as nonlinear Bayesian filtering: a new perspective

Autores: Filip Tronarp, Hans Kersting, Simo Särkkä, Philipp Hennig
Publicado en: Statistics and Computing, Edición 29/6, 2019, Página(s) 1297-1315, ISSN 0960-3174
Editor: Kluwer Academic Publishers
DOI: 10.1007/s11222-019-09900-1

Probabilistic linear solvers: a unifying view

Autores: Simon Bartels, Jon Cockayne, Ilse C. F. Ipsen, Philipp Hennig
Publicado en: Statistics and Computing, Edición 29/6, 2019, Página(s) 1249-1263, ISSN 0960-3174
Editor: Kluwer Academic Publishers
DOI: 10.1007/s11222-019-09897-7

Probabilistic Numerics: Computation as Machine Learning

Autores: Philipp Hennig, Michael A. Osborne, Hans P. Kersting
Publicado en: 2022, ISBN 9781316681411
Editor: Cambridge University Press
DOI: 10.1017/9781316681411

Publicaciones

Compartir esta página

Descargar