Zobrazeno 1 - 10
of 64
pro vyhledávání: '"Schlkopf, Bernhard"'
Publikováno v:
Advances in Neural Information Processing Systems 34
Many reinforcement learning (RL) environments consist of independent entities that interact sparsely. In such environments, RL agents have only limited influence over other entities in any particular situation. Our idea in this work is that learning
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::25a7316a8fad0a649b23b2f6de00d27c
https://proceedings.neurips.cc/paper_files/paper/2021/hash/c1722a7941d61aad6e651a35b65a9c3e-Abstract.html
https://proceedings.neurips.cc/paper_files/paper/2021/hash/c1722a7941d61aad6e651a35b65a9c3e-Abstract.html
Autor:
Mambelli, Davide, Tr��uble, Frederik, Bauer, Stefan, Sch��lkopf, Bernhard, Locatello, Francesco
Although reinforcement learning has seen remarkable progress over the last years, solving robust dexterous object-manipulation tasks in multi-object settings remains a challenge. In this paper, we focus on models that can learn manipulation tasks in
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::a6037d8c1668fc1b69f634d1b0c2686f
http://arxiv.org/abs/2201.13388
http://arxiv.org/abs/2201.13388
Model-free and model-based reinforcement learning are two ends of a spectrum. Learning a good policy without a dynamic model can be prohibitively expensive. Learning the dynamic model of a system can reduce the cost of learning the policy, but it can
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::dd94393fc1d3c513d06b993bad9932cc
http://arxiv.org/abs/2201.05830
http://arxiv.org/abs/2201.05830
Publikováno v:
International Conference on Learning Representations (ICLR 2022)
`Double descent' delineates the generalization behaviour of models depending on the regime they belong to: under- or over-parameterized. The current theoretical understanding behind the occurrence of this phenomenon is primarily based on linear and k
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::ff57759b84504b90ec95009f77832359
https://hdl.handle.net/20.500.11850/591657
https://hdl.handle.net/20.500.11850/591657
Distinguishing between cause and effect using time series observational data is a major challenge in many scientific fields. A new perspective has been provided based on the principle of Independence of Causal Mechanisms (ICM), leading to the Spectra
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::ebe98bc19b8edd037041b525813ce72d
http://arxiv.org/abs/2110.15595
http://arxiv.org/abs/2110.15595
Autor:
Makansi, Osama, von K��gelgen, Julius, Locatello, Francesco, Gehler, Peter, Janzing, Dominik, Brox, Thomas, Sch��lkopf, Bernhard
Predicting the future trajectory of a moving agent can be easy when the past trajectory continues smoothly but is challenging when complex interactions with other agents are involved. Recent deep learning approaches for trajectory prediction show pro
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::2dee1fd369d6f3107918a1fd95e6203e
http://arxiv.org/abs/2110.05304
http://arxiv.org/abs/2110.05304
Disentanglement is hypothesized to be beneficial towards a number of downstream tasks. However, a common assumption in learning disentangled representations is that the data generative factors are statistically independent. As current methods are alm
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::44e2077fba0c30d5b1bcae247426abaa
http://arxiv.org/abs/2110.03628
http://arxiv.org/abs/2110.03628
Autor:
Scherrer, Nino, Bilaniuk, Olexa, Annadani, Yashas, Goyal, Anirudh, Schwab, Patrick, Sch��lkopf, Bernhard, Mozer, Michael C., Bengio, Yoshua, Bauer, Stefan, Ke, Nan Rosemary
Discovering causal structures from data is a challenging inference problem of fundamental importance in all areas of science. The appealing properties of neural networks have recently led to a surge of interest in differentiable neural network-based
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::34681097fc116fdb72304a160742f46d
http://arxiv.org/abs/2109.02429
http://arxiv.org/abs/2109.02429
Source-free domain adaptation (SFDA) aims to adapt a model trained on labelled data in a source domain to unlabelled data in a target domain without access to the source-domain data during adaptation. Existing methods for SFDA leverage entropy-minimi
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::291c8f6b4d9a23b8db39077fdd919e00
http://arxiv.org/abs/2107.05446
http://arxiv.org/abs/2107.05446
It has been hypothesized that quantum computers may lend themselves well to applications in machine learning. In the present work, we analyze function classes defined via quantum kernels. Quantum computers offer the possibility to efficiently compute
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::2fe69840ad5f1d973e16c087152553db
http://arxiv.org/abs/2106.03747
http://arxiv.org/abs/2106.03747