You are currently offline. Some features of the site may not work correctly.

Corpus ID: 211010897

Interpreting a Penalty as the Influence of a Bayesian Prior

@article{Wolinski2020InterpretingAP,
title={Interpreting a Penalty as the Influence of a Bayesian Prior},
author={Pierre Wolinski and G. Charpiat and Y. Ollivier},
journal={ArXiv},
year={2020},
volume={abs/2002.00178}
}

In machine learning, it is common to optimize the parameters of a probabilistic model, modulated by a somewhat ad hoc regularization term that penalizes some values of the parameters. Regularization terms appear naturally in Variational Inference (VI), a tractable way to approximate Bayesian posteriors: the loss to optimize contains a Kullback--Leibler divergence term between the approximate posterior and a Bayesian prior. We fully characterize which regularizers can arise this way, and provide… Expand