Zobrazeno 1 - 10
of 246
pro vyhledávání: '"Lelarge, Marc"'
Data assimilation is a central problem in many geophysical applications, such as weather forecasting. It aims to estimate the state of a potentially large system, such as the atmosphere, from sparse observations, supplemented by prior physical knowle
Externí odkaz:
http://arxiv.org/abs/2406.15076
Exact Bayesian inference on state-space models~(SSM) is in general untractable, and unfortunately, basic Sequential Monte Carlo~(SMC) methods do not yield correct approximations for complex models. In this paper, we propose a mixed inference algorith
Externí odkaz:
http://arxiv.org/abs/2312.09860
Autor:
Blanke, Matthieu, Lelarge, Marc
Publikováno v:
The Twelfth International Conference on Learning Representations, ICLR 2024
Machine learning methods can be a valuable aid in the scientific process, but they need to face challenging settings where data come from inhomogeneous experimental conditions. Recent meta-learning methods have made significant progress in multi-task
Externí odkaz:
http://arxiv.org/abs/2312.00477
Autor:
Blanke, Matthieu, Lelarge, Marc
Model-based reinforcement learning is a powerful tool, but collecting data to fit an accurate model of the system can be costly. Exploring an unknown environment in a sample-efficient manner is hence of great importance. However, the complexity of dy
Externí odkaz:
http://arxiv.org/abs/2304.13426
In this paper, we present a new strategy to prove the convergence of deep learning architectures to a zero training (or even testing) loss by gradient flow. Our analysis is centered on the notion of Rayleigh quotients in order to prove Kurdyka-{\L}oj
Externí odkaz:
http://arxiv.org/abs/2301.08117
Autor:
Blanke, Matthieu, Lelarge, Marc
This work addresses the problem of exploration in an unknown environment. For linear dynamical systems, we use an experimental design framework and introduce an online greedy policy where the control maximizes the information of the next step. In a s
Externí odkaz:
http://arxiv.org/abs/2204.06375
For a very broad range of problems, recommendation algorithms have been increasingly used over the past decade. In most of these algorithms, the predictions are built upon user-item affinity scores which are obtained from high-dimensional embeddings
Externí odkaz:
http://arxiv.org/abs/2203.10107
Publikováno v:
Ann. Appl. Probab. 34 (3) 2799 - 2843, June 2024
Motivated by alignment of correlated sparse random graphs, we introduce a hypothesis testing problem of deciding whether or not two random trees are correlated. We obtain sufficient conditions under which this testing is impossible or feasible. We pr
Externí odkaz:
http://arxiv.org/abs/2107.07623
Publikováno v:
Proceedings of Thirty Fourth Conference on Learning Theory, PMLR 134:2080-2102, 2021
Random graph alignment refers to recovering the underlying vertex correspondence between two random graphs with correlated edges. This can be viewed as an average-case and noisy version of the well-known graph isomorphism problem. For the correlated
Externí odkaz:
http://arxiv.org/abs/2102.02685
Scarcity of training data for task-oriented dialogue systems is a well known problem that is usually tackled with costly and time-consuming manual data annotation. An alternative solution is to rely on automatic text generation which, although less a
Externí odkaz:
http://arxiv.org/abs/2011.02143