Probabilistic Automated Numerical Analysis in Machine learning and Artificial intelligence

Calibrated Adaptive Probabilistic ODE Solvers

Autorzy: Bosch, Nathanael; Hennig, Philipp; Tronarp, Filip
Opublikowane w: International Conference on Artificial Intelligence and Statistics, Numer 130, 2021, Strona(/y) 3466-3474
Wydawca: PMLR
DOI: 10.48550/arxiv.2012.08202

Counterfactual mean embeddings

Autorzy: Krikamol Muandet, Motonobu Kanagawa, Sorawit Saengkyongam, Sanparith Marukatat
Opublikowane w: The Journal of Machine Learning Research, Numer 22 (1), 2021, Strona(/y) 7322-7392
Wydawca: The Journal of Machine Learning Research

BackPACK: Packing more into backprop

Autorzy: Dangel, Felix; Kunstner, Frederik; Hennig, Philipp
Opublikowane w: ICLR, Numer 8, 2020
Wydawca: ICLR

High-Dimensional Gaussian Process Inference with Derivatives

Autorzy: de Roos, Filip; Gessner, Alexandra; Hennig, Philipp
Opublikowane w: International Conference on Machine Learning, Numer 139, 2021, Strona(/y) 2535-2545
Wydawca: PMLR

Probabilistic ODE solutions in millions of dimensions

Autorzy: Nicholas Krämer, Nathanael Bosch, Jonathan Schmidt, Philipp Hennig
Opublikowane w: International Conference on Machine Learning, Numer 162, 2022, Strona(/y) 11634-11649
Wydawca: PMLR

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

Autorzy: Nishiyama, Yu; Kanagawa, Motonobu; Gretton, Arthur; Fukumizu, Kenji
Opublikowane w: Machine Learning, Numer 109, 2020, Strona(/y) 939–972
Wydawca: Springer Link

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

Autorzy: Nicholas Krämer, Jonathan Schmidt, Philipp Hennig
Opublikowane w: Proceedings of The 25th International Conference on Artificial Intelligence and Statistics, Numer 151, 2022, Strona(/y) 625-639
Wydawca: PMLR

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

Autorzy: Kristiadi, Agustinus; Hein, Matthias; Hennig, Philipp
Opublikowane w: Advances in Neural Information Processing Systems, Numer 34, 2021, Strona(/y) 18789-18800
Wydawca: Curran Associates Inc
DOI: 10.48550/arxiv.2010.02709

Integrals over Gaussians under Linear Domain Constraints

Autorzy: Gessner, Alexandra; Kanjilal, Oindrila; Hennig, Philipp
Opublikowane w: Proceedings of Machine Learning Research, 2020, Strona(/y) 2764-2774
Wydawca: PMLR

Convergence Guarantees for Adaptive Bayesian Quadrature Methods

Autorzy: Kanagawa, Motonobu; Hennig, Philipp
Opublikowane w: Advances in Neural Information Processing Systems (NeurIPS 2019), Numer 32, 2019
Wydawca: Advances in Neural Information Processing Systems 32 (NeurIPS 2019)
DOI: 10.48550/arxiv.1905.10271

Descending through a Crowded Valley - Benchmarking Deep Learning Optimizers

Autorzy: Schmidt, Robin M.; Schneider, Frank; Hennig, Philipp
Opublikowane w: International Conference on Machine Learning, Numer 139, 2021, Strona(/y) 9367-9376
Wydawca: PMLR
DOI: 10.48550/arxiv.2007.01547

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

Autorzy: Kajihara, Takafumi; Kanagawa, Motonobu; Nakaguchi, Yuuki; Khandelwal, Kanishka; Fukumiziu, Kenji
Opublikowane w: Machine Learning, Numer 109, 2020, Strona(/y) 939–972
Wydawca: Springer
DOI: 10.48550/arxiv.1902.02517

Fenrir: Physics-Enhanced Regression for Initial Value Problems

Autorzy: Filip Tronarp, Nathanael Bosch, Philipp Hennig
Opublikowane w: International Conference on Machine Learning, Numer 162, 2022, Strona(/y) 21776--21794
Wydawca: PMLR

DeepOBS: A Deep Learning Optimizer Benchmark Suite

Autorzy: Schneider, Frank; Balles, Lukas; Hennig, Philipp
Opublikowane w: International Conference on Learning Representations, 2019
Wydawca: International Conference on Learning Representations
DOI: 10.48550/arxiv.1903.05499

The Geometry of Sign Gradient Descent

Autorzy: Balles, Lukas; Pedregosa, Fabian; Le Roux, Nicolas
Opublikowane w: ICLR 2020, 2020
Wydawca: ICLR

Fast Predictive Uncertainty for Classification with Bayesian Deep Networks

Autorzy: Hobbhahn, Marius; Kristiadi, Agustinus; Hennig, Philipp
Opublikowane w: Uncertainty in Artificial Intelligence, Numer 180, 2022, Strona(/y) 822-832
Wydawca: PMLR
DOI: 10.48550/arxiv.2003.01227

Preconditioning for scalable Gaussian process hyperparameter optimization

Autorzy: Jonathan Wenger, Geoff Pleiss, Philipp Hennig, John Cunningham, Jacob Gardner
Opublikowane w: International Conference on Machine Learning, Numer 162, 2022, Strona(/y) 23751-23780
Wydawca: PMLR

Learnable Uncertainty under Laplace Approximations

Autorzy: Kristiadi, Agustinus; Hein, Matthias; Hennig, Philipp
Opublikowane w: Proceedings of the Thirty-Seventh Conference on Uncertainty in Artificial Intelligence, Numer 161, 2021, Strona(/y) 344-353
Wydawca: PMLR
DOI: 10.48550/arxiv.2010.02720

Laplace Redux -- Effortless Bayesian Deep Learning

Autorzy: Daxberger, Erik; Kristiadi, Agustinus; Immer, Alexander; Eschenhagen, Runa; Bauer, Matthias; Hennig, Philipp
Opublikowane w: Advances in Neural Information Processing Systems (NeurIPS), Numer 34, 2021, Strona(/y) 20089-20103
Wydawca: 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

Autorzy: Jonathan Oesterle, Nicholas Krämer, Philipp Hennig, Philipp Berens
Opublikowane w: Journal of Computational Neuroscience, Numer 50 (4), 2022, Strona(/y) 485-503
Wydawca: Springer US
DOI: 10.1007/s10827-022-00827-7

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

Autorzy: Kristiadi, Agustinus; Hein, Matthias; Hennig, Philipp
Opublikowane w: ICML, Numer 4, 2019
Wydawca: ICML

Probabilistic Linear Solvers for Machine Learning

Autorzy: Wenger, Jonathan; Hennig, Philipp
Opublikowane w: Advances in Neural Information Processing Systems, Numer 33, 2020, Strona(/y) 6731 - 6742
Wydawca: Curran Associate Inc.
DOI: 10.48550/arxiv.2010.09691

Limitations of the empirical Fisher approximation for natural gradient descent

Autorzy: Kunstner, Frederik; Hennig, Philipp; Balles, Lukas
Opublikowane w: Advances in Neural Information Processing Systems 32, Numer 32, 2019, Strona(/y) {4158--4169
Wydawca: Curran Associates, Inc.

Pick-and-mix information operators for probabilistic ODE solvers

Autorzy: Nathanael Bosch, Filip Tronarp, Philipp Hennig
Opublikowane w: International Conference on Artificial Intelligence and Statistics, Numer 151, 2022, Strona(/y) 10015-10027
Wydawca: PMLR

A Fourier State Space Model for Bayesian ODE Filters

Autorzy: Kersting, Hans; Mahsereci, Maren
Opublikowane w: Workshop on Invertible Neural Networks, Normalizing Flows, and Explicit Likelihood Models, ICML, 2020
Wydawca: ICML

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

Autorzy: Schneider, Frank; Dangel, Felix; Hennig, Philipp
Opublikowane w: Advances in Neural Information Processing Systems, Numer 34, 2021, Strona(/y) 20825-20837
Wydawca: Curran Associates Inc.
DOI: 10.48550/arxiv.2102.06604

Kernel Recursive ABC: Point Estimation with Intractable Likelihood

Autorzy: Takafumi Kajihara, Motonobu Kanagawa, Keisuke Yamazaki, and Kenji Fukumizu
Opublikowane w: Proceedings of the International Conference on Machine Learning, Numer 35, 2018, Strona(/y) 2400-2409
Wydawca: PMLR (Proceedings of Machine Learning Research

Convergence Guarantees for Adaptive Bayesian Quadrature Methods

Autorzy: Motonobu Kanagawa, Philipp Hennig
Opublikowane w: Advances in Neural Information Processing Systems (NeurIPS), Numer 32, 2019, Strona(/y) 6234--6245
Wydawca: Curran Associates, Inc.

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

Autorzy: Lukas Balles, Philipp Hennig
Opublikowane w: Proceedings of the 35th International Conference on Machine Learning (ICML), Numer 35, 2018, Strona(/y) 404--413
Wydawca: PMLR

DeepOBS: A Deep Learning Optimizer Benchmark Suite

Autorzy: Schneider, Frank; Balles, Lukas; Hennig, Philipp
Opublikowane w: International Conference on Learning Representations (ICLR), Numer 7, 2019
Wydawca: ICLR

Limitations of the empirical Fisher approximation for natural gradient descent

Autorzy: Frederik Kunstner, Philipp Hennig, Lukas Balles
Opublikowane w: Advances in Neural Information Processing Systems (NeurIPS), Numer 32, 2019, Strona(/y) 4158--4169
Wydawca: Curran Associates, Inc.

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

Autorzy: Filip de Roos, Philipp Hennig
Opublikowane w: Proceedings of the 22nd International Conference on Artificial Intelligence and Statistics (AISTATS), Numer 22, 2019, Strona(/y) 1448--1457
Wydawca: PMLR

Fast and Robust Shortest Paths on Manifolds Learned from Data

Autorzy: Georgios Arvanitidis, Soren Hauberg, Philipp Hennig, Michael Schober
Opublikowane w: Proceedings of the 22nd International Conference on Artificial Intelligence and Statistics (AISTATS), Numer 22, 2019, Strona(/y) 1506--1515
Wydawca: JMLR

Active Multi-Information Source Bayesian Quadrature

Autorzy: Alexandra Gessner, Javier Gonzalez, Maren Mahsereci
Opublikowane w: Conference on Uncertainty in Artificial Intelligence (UAI), Numer 35, 2019
Wydawca: UAI

Convergence rates of Gaussian ODE filters

Autorzy: Hans Kersting; Timothy Sullivan; Philipp Hennig
Opublikowane w: Statistics and computing, Numer 30 (6), 2020, Strona(/y) 1791-1816
Wydawca: Springer US
DOI: 10.1007/s11222-020-09972-4

Being a Bit Frequentist Improves Bayesian Neural Networks

Autorzy: Kristiadi, Agustinus; Hein, Matthias; Hennig, Philipp
Opublikowane w: International Conference on Artificial Intelligence and Statistics, Numer 151, 2022, Strona(/y) 529-545
Wydawca: PMLR
DOI: 10.48550/arxiv.2106.10065

Linear-Time Probabilistic Solutions of Boundary Value Problems

Autorzy: Krämer, Nicholas; Hennig, Philipp
Opublikowane w: Advances in Neural Information Processing Systems 34 (NeurIPS 2021), Numer 34, 2021, Strona(/y) 11160-11171
Wydawca: Advances in Neural Information Processing Systems 34 (NeurIPS 2021)

Probabilistic DAG Search

Autorzy: Grosse, Julia; Zhang, Cheng; Hennig, Philipp
Opublikowane w: Uncertainty in Artificial Intelligence, Numer 161, 2021, Strona(/y) 1424-1433
Wydawca: PMLR
DOI: 10.48550/arxiv.2106.08717

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

Autorzy: Schmidt, Jonathan; Krämer, Nicholas; Hennig, Philipp
Opublikowane w: Advances in Neural Information Processing Systems 34 (NeurIPS 2021), Numer 34, 2021, Strona(/y) 12374-12385
Wydawca: Advances in Neural Information Processing Systems 34 (NeurIPS 2021)
DOI: 10.48550/arxiv.2103.10153

ProbNum: Probabilistic Numerics in Python

Autorzy: 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
Opublikowane w: 2021
Wydawca: arXiv
DOI: 10.48550/arxiv.2112.02100

Bayesian Quadrature on Riemannian Data Manifolds

Autorzy: Fröhlich, Christian; Gessner, Alexandra; Hennig, Philipp; Schölkopf, Bernhard; Arvanitidis, Georgios
Opublikowane w: International Conference on Machine Learning, Numer 139, 2021, Strona(/y) 3459-3468
Wydawca: PMLR
DOI: 10.48550/arxiv.2102.06645

Integrals over Gaussians under Linear Domain Constraints

Autorzy: Gessner, Alexandra; Kanjilal, Oindrila; Hennig, Philipp
Opublikowane w: International Conference on Artificial Intelligence and Statistics. Proceedings of Machine Learning Research (PMLR), Numer 1, 2020
Wydawca: MLR Press
DOI: 10.48550/arxiv.1910.09328

Modular Block-diagonal Curvature Approximations for Feedforward Architectures

Autorzy: Dangel, Felix; Harmeling, Stefan; Hennig, Philipp
Opublikowane w: Proceedings of the Twenty Third International Conference on Artificial Intelligence and Statistics, Numer 2, 2020, Strona(/y) 799-808
Wydawca: PMLR

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

Autorzy: Kersting, Hans; Krämer, Nicholas; Schiegg, Martin; Daniel, Christian; Tiemann, Michael; Hennig, Philipp
Opublikowane w: International Conference on Machine Learning (ICML), Numer 11, 2020
Wydawca: ICML

Resnet after all: Neural odes and their numerical solution

Autorzy: Katharina Ott, Prateek Katiyar, Philipp Hennig, Michael Tiemann
Opublikowane w: International Conference on Learning Representations, 2021
Wydawca: International Conference on Learning Representations

Conjugate Gradients for Kernel Machines

Autorzy: Bartels, Simon; Hennig, Philipp
Opublikowane w: Journal of Machine Learning Research, 2020, Strona(/y) 1-42
Wydawca: Journal of Machine Learning Research

Bayesian ODE solvers: the maximum a posteriori estimate

Autorzy: Filip Tronarp; Simo Särkkä; Philipp Hennig
Opublikowane w: Statistics and Computing, 31(3), Numer 31, 2021, ISSN 0960-3174
Wydawca: Kluwer Academic Publishers
DOI: 10.1007/s11222-021-09993-7

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

Autorzy: Motonobu Kanagawa, Bharath K. Sriperumbudur, Kenji Fukumizu
Opublikowane w: Foundations of Computational Mathematics, 2019, Strona(/y) 1-40, ISSN 1615-3375
Wydawca: Springer Verlag
DOI: 10.1007/s10208-018-09407-7

On the positivity and magnitudes of Bayesian quadrature weights

Autorzy: Toni Karvonen, Motonobu Kanagawa, Simo Särkkä
Opublikowane w: Statistics and Computing, Numer 29/6, 2019, Strona(/y) 1317-1333, ISSN 0960-3174
Wydawca: Kluwer Academic Publishers
DOI: 10.1007/s11222-019-09901-0

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

Autorzy: Filip Tronarp, Hans Kersting, Simo Särkkä, Philipp Hennig
Opublikowane w: Statistics and Computing, Numer 29/6, 2019, Strona(/y) 1297-1315, ISSN 0960-3174
Wydawca: Kluwer Academic Publishers
DOI: 10.1007/s11222-019-09900-1

Probabilistic linear solvers: a unifying view

Autorzy: Simon Bartels, Jon Cockayne, Ilse C. F. Ipsen, Philipp Hennig
Opublikowane w: Statistics and Computing, Numer 29/6, 2019, Strona(/y) 1249-1263, ISSN 0960-3174
Wydawca: Kluwer Academic Publishers
DOI: 10.1007/s11222-019-09897-7

Probabilistic Numerics: Computation as Machine Learning

Autorzy: Philipp Hennig, Michael A. Osborne, Hans P. Kersting
Opublikowane w: 2022, ISBN 9781316681411
Wydawca: Cambridge University Press
DOI: 10.1017/9781316681411

Publikacje

Udostępnij tę stronę

Pobierz