Zobrazeno 1 - 2
of 2
pro vyhledávání: '"Casado, Ioar"'
We introduce a new PAC-Bayes oracle bound for unbounded losses that extends Cram\'er-Chernoff bounds to the PAC-Bayesian setting. The proof technique relies on controlling the tails of certain random variables involving the Cram\'er transform of the
Externí odkaz:
http://arxiv.org/abs/2401.01148
Autor:
Casado, Ioar, Pérez, Aritz
In recent years, we have seen a handful of work on inference algorithms over non-stationary data streams. Given their flexibility, Bayesian non-parametric models are a good candidate for these scenarios. However, reliable streaming inference under th
Externí odkaz:
http://arxiv.org/abs/2210.06872