Efficient algorithms for sustainable machine learning

Información del proyecto

SLING

Identificador del acuerdo de subvención: 819789

DOI

10.3030/819789

Proyecto cerrado

Fecha de la firma de la CE 21 Marzo 2019

Fecha de inicio 1 Noviembre 2019

Fecha de finalización 31 Octubre 2025

Financiado con arreglo a

• EXCELLENT SCIENCE - European Research Council (ERC)

Coste total

€ 1 977 500,00

Aportación de la UE

€ 1 977 500,00

Coordinado por

UNIVERSITA DEGLI STUDI DI GENOVA
Italy

CORDIS proporciona enlaces a los documentos públicos y las publicaciones de los proyectos de los programas marco HORIZONTE.

Los enlaces a los documentos y las publicaciones de los proyectos del Séptimo Programa Marco, así como los enlaces a algunos tipos de resultados específicos, como conjuntos de datos y «software», se obtienen dinámicamente de OpenAIRE .

Publicaciones

Conference proceedings (25)

Fast object segmentation learning with kernel-based methods for robotics

Autores: Federico Ceola, Elisa Maiettini, Giulia Pasquale, Lorenzo Rosasco, Lorenzo Natale
Publicado en: 2021 IEEE International Conference on Robotics and Automation (ICRA), 2021, Página(s) 13581-13588
Editor: IEEE

An optimal structured zeroth-order algorithm for non-smooth optimization

Autores: Marco Rando, Cesare Molinari, Lorenzo Rosasco, Silvia Villa
Publicado en: Advances in Neural Information Processing Systems, Edición 36, 2024
Editor: Advances in Neural Information Processing Systems

Efficient Unsupervised Learning for Plankton Images

Autores: P. D. Alfano; M. Rando; M. Letizia; F. Odone; L. Rosasco; V. P. Pastore
Publicado en: 2022 26th International Conference on Pattern Recognition (ICPR), 2022
Editor: IEEE
DOI: 10.1109/icpr56361.2022.9956360

Snacks: a fast large-scale kernel SVM solver

Autores: Tanji, Sofiane; Vecchia, Andrea Della; Glineur, François; Villa, Silvia; 2023 European Control Conference (ECC)
Publicado en: 2023 European Control Conference (ECC), Edición 4, 2023, Página(s) 1-6
Editor: IEEE
DOI: 10.23919/ecc57647.2023.10178323

Mean nyström embeddings for adaptive compressive learning

Autores: Antoine Chatalic, Luigi Carratino, Ernesto De Vito, Lorenzo Rosasco
Publicado en: International Conference on Artificial Intelligence and Statistics, 2022, Página(s) 9869-9889
Editor: PMLR

Decentralised learning with random features and distributed gradient descent

Autores: Dominic Richards, Patrick Rebeschini, Lorenzo Rosasco
Publicado en: Proceedings of Machine Learning Research, 2020
Editor: JMLR, Inc. and Microtome Publishing

Iterative regularization for convex regularizers

Autores: Cesare Molinari, Mathurin Massias, Lorenzo Rosasco, Silvia Villa
Publicado en: Proceedings of Machine Learning Research, Edición 130, 2021, Página(s) 1684-1692
Editor: JMLR, Inc. and Microtome Publishing

Ada-BKB: Scalable Gaussian Process Optimization on Continuous Domains by Adaptive Discretization

Autores: Rando, Marco; Carratino, Luigi; Villa, Silvia; Rosasco, Lorenzo
Publicado en: International Conference on Artificial Intelligence and Statistics, Edición 9, 2022, Página(s) 320-7348
Editor: PLMR
DOI: 10.48550/arxiv.2106.08598

Gain with no pain: Efficiency of kernel-PCA by Nyström sampling

Autores: Nicholas Sterge, Bharath Sriperumbudur, Lorenzo Rosasco, Alessandro Rudi
Publicado en: International Conference on Artificial Intelligence and Statistics, 2020, Página(s) 3642-3652
Editor: AISTATS

Nyström Kernel Mean Embeddings

Autores: Antoine Chatalic, Nicolas Schreuder, Lorenzo Rosasco, Alessandro Rudi
Publicado en: International Conference on Machine Learning, 2022
Editor: PMLR

Estimating Koopman operators with sketching to provably learn large scale dynamical systems

Autores: Meanti, Giacomo; Chatalic, Antoine; Kostic, Vladimir R.; Novelli, Pietro; Pontil, Massimiliano; Rosasco, Lorenzo
Publicado en: Proceedings of the 37th International Conference on Neural Information Processing Systems, Edición 12, 2023, Página(s) 77242–77276
Editor: Curran Associates
DOI: 10.48550/arxiv.2306.04520

Asymptotics of Ridge(less) Regression under General Source Condition

Autores: Dominic Richards, Jaouad Mourtada, Lorenzo Rosasco
Publicado en: Proceedings of Machine Learning Research, Edición 130, 2021, Página(s) 3889-3897
Editor: JMLR, Inc. and Microtome Publishing

Efficient kernel methods for model-independent new physics searches

Autores: Marco Letizia, Gianvito Losapio, Marco Rando, Gaia Grosso, Lorenzo Rosasco
Publicado en: NeurIPS 2022 workshop: Machine Learning and the Physical Sciences, 2021
Editor: NeurIPS
DOI: 10.5281/zenodo.5536345

Multiclass learning with margin: exponential rates with no bias-variance trade-off

Autores: Stefano Vigogna, Giacomo Meanti, Ernesto De Vito, Lorenzo Rosasco
Publicado en: International Conference on Machine Learning, 2022
Editor: Association for Computing Machinery

Hyperbolic manifold regression

Autores: Gian Maria Marconi, Carlo Ciliberto, Lorenzo Rosasco
Publicado en: Proceedings of Machine Learning Research, 2020
Editor: JMLR, Inc. and Microtome Publishing

Regularized ERM on random subspaces

Autores: Andrea Della Vecchia, Jaouad Mourtada, Ernesto De Vito, Lorenzo Rosasco
Publicado en: Proceedings of Machine Learning Research, Edición 130, 2021, Página(s) 4006-4014
Editor: JMLR, Inc. and Microtome Publishing

Fair learning with Wasserstein barycenters for non-decomposable performance measures

Autores: Solenne Gaucher, Nicolas Schreuder, Evgenii Chzhen
Publicado en: International Conference on Artificial Intelligence and Statistics, Edición 2436-2459, 2023
Editor: PMLR

A fast and flexible machine learning approach to data quality monitoring

Autores: Grosso, Gaia; Lai, Nicolò; Letizia, Marco; Pazzini, Jacopo; Rando, Marco; Wulzer, Andrea; Zanetti, Marco
Publicado en: NeurIPS 2022 workshop: Machine Learning and the Physical Sciences, Edición 1, 2023
Editor: NeurIPS
DOI: 10.48550/arxiv.2301.08917

Kernel methods through the roof: Handling billions of points efficiently

Autores: Giacomo Meanti, Luigi Carratino, Lorenzo Rosasco, Alessandro Rudi
Publicado en: Advances in Neural Information Processing Systems, 2020
Editor: Neural information processing systems foundation

Efficient Hyperparameter Tuning for Large Scale Kernel Ridge Regression

Autores: Giacomo Meanti, Luigi Carratino, Ernesto De Vito, Lorenzo Rosasco
Publicado en: International Conference on Artificial Intelligence and Statistics, 2022, Página(s) 6554-6572
Editor: PMLR

Regularization Properties of Dual Subgradient Flow

Autores: Apidopoulos, Vassilis; Molinari, Cesare; Rosasco, Lorenzo; Villa, Silvia
Publicado en: 2023 European Control Conference ({ECC}), Edición 4, 2023, Página(s) 1-8
Editor: IEEE
DOI: 10.23919/ecc57647.2023.10178128

Near-linear time gaussian process optimization with adaptive batching and resparsification

Autores: Daniele Calandriello, Luigi Carratino, Alessandro Lazaric, Michal Valko, Lorenzo Rosasco
Publicado en: Proceedings of Machine Learning Research, Edición 119, 2020, Página(s) 1295-1305
Editor: JMLR, Inc. and Microtome Publishing

Park: Sound and efficient kernel ridge regression by feature space partitions

Autores: Luigi Carratino, Stefano Vigogna, Daniele Calandriello, Lorenzo Rosasco
Publicado en: Advances in Neural Information Processing Systems, Edición 34, 2021, Página(s) 6430-6441
Editor: NeuIPS

Heteroscedastic Gaussian Processes and Random Features: Scalable Motion Primitives with Guarantees

Autores: Caldarelli, E; Chatalic, A; Colome, A; Rosasco, L; Torras, C
Publicado en: 7th Conference on Robot Learning (CoRL 2023), 2023
Editor: MLR Press

Anderson acceleration of coordinate descent

Autores: Quentin Bertrand, Mathurin Massias
Publicado en: Proceedings of Machine Learning Research, Edición 130, 2021, Página(s) 1288-1296
Editor: JMLR, Inc. and Microtome Publishing

Peer reviewed articles (28)

Fast approximation of orthogonal matrices and application to PCA

Autores: Christian Rusu, Lorenzo Rosasco
Publicado en: Signal processing, Volume 194, 2022, ISSN 0165-1684
Editor: Elsevier BV
DOI: 10.1016/j.sigpro.2021.108451

On the emergence of whole-body strategies from humanoid robot push-recovery learning

Autores: Diego Ferigo, Raffaello Camoriano, Paolo Maria Viceconte, Daniele Calandriello, Silvio Traversaro, Lorenzo Rosasco, Daniele Pucci
Publicado en: IEEE Robotics and Automation Letters, 2021, ISSN 2377-3766
Editor: Institute of Electrical and Electronics Engineers Inc.

Multi-scale vector quantization with reconstruction trees

Autores: Enrico Cecini, Ernesto De Vito, Lorenzo Rosasco
Publicado en: Information and Inference: A Journal of the IMA, 2021, Página(s) 955-986, ISSN 2049-8772
Editor: Oxford University Press

A General Framework for Consistent Structured Prediction with Implicit Loss Embeddings

Autores: Carlo Ciliberto, Lorenzo Rosasco, Alessandro Rudi
Publicado en: J. Mach. Learn. Res., 2020, ISSN 1532-4435
Editor: MIT Press

Convergence rates for the Heavy-Ball continuous dynamics for non-convex optimization, under Polyak-\L ojasiewicz condition

Autores: Apidopoulos, Vassilis; Ginatta, Nicolò; Villa, Silvia
Publicado en: Journal of Global Optimization, Edición 8, 2022, ISSN 0925-5001
Editor: Kluwer Academic Publishers

An elementary analysis of ridge regression with random design

Autores: Jaouad Mourtada, Lorenzo Rosasco
Publicado en: Comptes Rendus. Mathématique, 2022, ISSN 1778-3569
Editor: Academie des sciences

Accelerated Iterative Regularization via Dual Diagonal Descent

Autores: Luca Calatroni, Guillaume Garrigos, Lorenzo Rosasco, Silvia Villa
Publicado en: SIAM Journal on Optimization, 2021, ISSN 1052-6234
Editor: Society for Industrial and Applied Mathematics
DOI: 10.1137/19m1308888

Construction and Monte Carlo Estimation of Wavelet Frames Generated by a Reproducing Kernel

Autores: Ernesto De Vito, Zeljko Kereta, Valeriya Naumova, Lorenzo Rosasco, Stefano Vigogna
Publicado en: Journal of Fourier Analysis and Applications, Edición 27/2, 2021, ISSN 1069-5869
Editor: Birkhaeuser
DOI: 10.1007/s00041-021-09835-0

Iterative regularization for low complexity regularizers

Autores: Molinari, Cesare; Massias, Mathurin; Rosasco, Lorenzo; Villa, Silvia
Publicado en: Numerische Mathematik, Edición Volume 156, 2024, Página(s) 641–689, ISSN 0029-599X
Editor: Springer Verlag
DOI: 10.1007/s00211-023-01390-8

Convergence of the forward-backward algorithm: beyond the worst-case with the help of geometry

Autores: Guillaume Garrigos, Lorenzo Rosasco, Silvia Villa
Publicado en: Mathematical Programming, 2022, Página(s) 1-60, ISSN 0025-5610
Editor: Springer Verlag

Goodness of fit by Neyman-Pearson testing

Autores: Grosso, Gaia; Letizia, Marco; Pierini, Maurizio; Wulzer, Andrea
Publicado en: SciPost Physics, Edición Volume 16, Number 5, 2024, ISSN 2542-4653
Editor: SciPost Foundation
DOI: 10.48550/arxiv.2305.14137

Implicit regularization with strongly convex bias: stability and acceleration

Autores: Silvia Villa, Simon Matet, Bằng Công Vũ, Lorenzo Rosasco
Publicado en: Analysis and Applications, 2022, ISSN 0219-5305
Editor: World Scientific

Learning to Avoid Obstacles With Minimal Intervention Control

Autores: Anqing Duan, Raffaello Camoriano, Diego Ferigo, Yanlong Huang, Daniele Calandriello, Lorenzo Rosasco, Daniele Pucci
Publicado en: Frontiers in Robotics and AI, Edición 7, 2020, ISSN 2296-9144
Editor: Frontiers Media S.A.
DOI: 10.3389/frobt.2020.00060

Understanding neural networks with reproducing kernel Banach spaces

Autores: Ernesto De Vito, Lorenzo Rosasco
Publicado en: Applied and Computational Harmonic Analysis, Edición 62, 2023, Página(s) 194-236, ISSN 1096-603X
Editor: Academic Press Inc.
DOI: 10.1016/j.acha.2022.08.006

Reproducing kernel Hilbert spaces on manifolds: Sobolev and diffusion spaces

Autores: Ernesto De Vito, Nicole Mücke, Lorenzo Rosasco
Publicado en: Analysis and Applications, Edición 19-03, 2021, Página(s) 363-396, ISSN 1793-6861
Editor: World Scientific Publishing Co. Pte Ltd
DOI: 10.1142/s0219530520400114

Structured prediction for CRiSP inverse kinematics learning with misspecified robot models

Autores: Gian Maria Marconi, Raffaello Camoriano, Lorenzo Rosasco, Carlo Ciliberto
Publicado en: IEEE Robotics and Automation Letters, 2021, ISSN 2377-3766
Editor: Institute of Electrical and Electronics Engineers Inc.

Faster kriging: Facing high-dimensional simulators

Autores: Xuefei Lu, Alessandro Rudi, Emanuele Borgonovo, Lorenzo Rosasco
Publicado en: Operations Research, 2020, ISSN 0030-364X
Editor: Institute for Operations Research and the Management Sciences

On-line object detection: a robotics challenge

Autores: Elisa Maiettini, Giulia Pasquale, Lorenzo Rosasco, Lorenzo Natale
Publicado en: Autonomous Robots, 2020, ISSN 1573-7527
Editor: Springer Netherlands

Top-tuning: A study on transfer learning for an efficient alternative to fine tuning for image classification with fast kernel methods

Autores: Alfano P. D.; Pastore V. P.; Rosasco L.; Odone F.
Publicado en: Image and Vision Computing, Edición Volume 142, 2024, ISSN 0262-8856
Editor: Elsevier BV
DOI: 10.1016/j.imavis.2023.104894

Adherent: Learning human-like trajectory generators for whole-body control of humanoid robots

Autores: Paolo Maria Viceconte, Raffaello Camoriano, Giulio Romualdi, Diego Ferigo, Stefano Dafarra, Silvio Traversaro, Giuseppe Oriolo, Lorenzo Rosasco, Daniele Pucci
Publicado en: IEEE Robotics and Automation Letters, Edición 7-2, 2022, Página(s) 2779-2786, ISSN 2377-3766
Editor: IEEE

Constructing fast approximate eigenspaces with application to the fast graph Fourier transforms

Autores: Cristian Rusu, Lorenzo Rosasco
Publicado en: IEEE Transactions on Signal Processing, 2021, Página(s) 1-1, ISSN 1053-587X
Editor: Institute of Electrical and Electronics Engineers
DOI: 10.1109/tsp.2021.3107629

Learning new physics efficiently with nonparametric methods

Autores: Marco Letizia; Gianvito Losapio; Marco Rando; Gaia Grosso; Andrea Wulzer; Maurizio Pierini; Marco Zanetti; Lorenzo Rosasco
Publicado en: European Physical Journal, Edición 6, 2022, ISSN 1434-6044
Editor: Springer Verlag
DOI: 10.1140/epjc/s10052-022-10830-y

Zeroth-order optimization with orthogonal random directions

Autores: David Kozak; Cesare Molinari; Lorenzo Rosasco; Luis Tenorio; Silvia Villa
Publicado en: Mathematical Programming, Edición 6, 2022, ISSN 0025-5610
Editor: Springer Verlag
DOI: 10.1007/s10107-022-01866-9

Convergence of an asynchronous block-coordinate forward-backward algorithm for convex composite optimization

Autores: Cheik Traoré; Saverio Salzo; Silvia Villa
Publicado en: Computational Optimization and Applications, Edición 86, 2023, Página(s) 303–344, ISSN 0926-6003
Editor: Kluwer Academic Publishers
DOI: 10.1007/s10589-023-00489-w

Physics informed machine learning for wind speed prediction

Autores: Daniele Lagomarsino-Oneto, Giacomo Meanti, Nicolò Pagliana, Alessandro Verri, Andrea Mazzino, Lorenzo Rosasco, Agnese Seminara
Publicado en: Energy, 2023, ISSN 0360-5442
Editor: Pergamon Press Ltd.

Learn Fast, Segment Well: Fast Object Segmentation Learning on the iCub Robot

Autores: Federico Ceola; Elisa Maiettini; Giulia Pasquale; Giacomo Meanti; Lorenzo Rosasco; Lorenzo Natale
Publicado en: IEEE Transactions on Robotics, Edición 2, 2022, Página(s) 3154-3172, ISSN 1552-3098
Editor: Institute of Electrical and Electronics Engineers
DOI: 10.1109/tro.2022.3164331

Fast iterative regularization by reusing data

Autores: Cristian Vega; Cesare Molinari; Lorenzo Rosasco; Silvia Villa
Publicado en: Journal of Inverse and Ill-posed Problems, Edición 3, 2023, ISSN 0928-0219
Editor: Walter de Gruyter GmbH & Co. KG
DOI: 10.1515/jiip-2023-0009

Fast kernel methods for data quality monitoring as a goodness-of-fit test

Autores: Gaia Grosso; Nicolò Lai; Marco Letizia; Jacopo Pazzini; Marco Rando; Lorenzo Rosasco; Andrea Wulzer; Marco Zanetti
Publicado en: Machine Learning: Science and Technology, 4 (3) 035029, Edición 6, 2023, ISSN 2632-2153
Editor: Institute of Physics Publishing
DOI: 10.1088/2632-2153/acebb7

Book chapters (1)

Regularization: From inverse problems to large-scale machine learning

Autores: Ernesto De Vito, Lorenzo Rosasco, Alessandro Rudi
Publicado en: 2021, Página(s) 245-296
Editor: Springer International Publishing AG