Rate-Distortion Analysis of Minimum Excess Risk in Bayesian Learning

Hassan Hafez-Kolahi, Behrad Moniri, Shohreh Kasaei, Mahdieh Soleymani Baghshah
Proceedings of the 38th International Conference on Machine Learning, PMLR 139:3998-4007, 2021.

Abstract

In parametric Bayesian learning, a prior is assumed on the parameter $W$ which determines the distribution of samples. In this setting, Minimum Excess Risk (MER) is defined as the difference between the minimum expected loss achievable when learning from data and the minimum expected loss that could be achieved if $W$ was observed. In this paper, we build upon and extend the recent results of (Xu & Raginsky, 2020) to analyze the MER in Bayesian learning and derive information-theoretic bounds on it. We formulate the problem as a (constrained) rate-distortion optimization and show how the solution can be bounded above and below by two other rate-distortion functions that are easier to study. The lower bound represents the minimum possible excess risk achievable by \emph{any} process using $R$ bits of information from the parameter $W$. For the upper bound, the optimization is further constrained to use $R$ bits from the training set, a setting which relates MER to information-theoretic bounds on the generalization gap in frequentist learning. We derive information-theoretic bounds on the difference between these upper and lower bounds and show that they can provide order-wise tight rates for MER under certain conditions. This analysis gives more insight into the information-theoretic nature of Bayesian learning as well as providing novel bounds.

Cite this Paper


BibTeX
@InProceedings{pmlr-v139-hafez-kolahi21a, title = {Rate-Distortion Analysis of Minimum Excess Risk in Bayesian Learning}, author = {Hafez-Kolahi, Hassan and Moniri, Behrad and Kasaei, Shohreh and Baghshah, Mahdieh Soleymani}, booktitle = {Proceedings of the 38th International Conference on Machine Learning}, pages = {3998--4007}, year = {2021}, editor = {Meila, Marina and Zhang, Tong}, volume = {139}, series = {Proceedings of Machine Learning Research}, month = {18--24 Jul}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v139/hafez-kolahi21a/hafez-kolahi21a.pdf}, url = {https://proceedings.mlr.press/v139/hafez-kolahi21a.html}, abstract = {In parametric Bayesian learning, a prior is assumed on the parameter $W$ which determines the distribution of samples. In this setting, Minimum Excess Risk (MER) is defined as the difference between the minimum expected loss achievable when learning from data and the minimum expected loss that could be achieved if $W$ was observed. In this paper, we build upon and extend the recent results of (Xu & Raginsky, 2020) to analyze the MER in Bayesian learning and derive information-theoretic bounds on it. We formulate the problem as a (constrained) rate-distortion optimization and show how the solution can be bounded above and below by two other rate-distortion functions that are easier to study. The lower bound represents the minimum possible excess risk achievable by \emph{any} process using $R$ bits of information from the parameter $W$. For the upper bound, the optimization is further constrained to use $R$ bits from the training set, a setting which relates MER to information-theoretic bounds on the generalization gap in frequentist learning. We derive information-theoretic bounds on the difference between these upper and lower bounds and show that they can provide order-wise tight rates for MER under certain conditions. This analysis gives more insight into the information-theoretic nature of Bayesian learning as well as providing novel bounds.} }
Endnote
%0 Conference Paper %T Rate-Distortion Analysis of Minimum Excess Risk in Bayesian Learning %A Hassan Hafez-Kolahi %A Behrad Moniri %A Shohreh Kasaei %A Mahdieh Soleymani Baghshah %B Proceedings of the 38th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2021 %E Marina Meila %E Tong Zhang %F pmlr-v139-hafez-kolahi21a %I PMLR %P 3998--4007 %U https://proceedings.mlr.press/v139/hafez-kolahi21a.html %V 139 %X In parametric Bayesian learning, a prior is assumed on the parameter $W$ which determines the distribution of samples. In this setting, Minimum Excess Risk (MER) is defined as the difference between the minimum expected loss achievable when learning from data and the minimum expected loss that could be achieved if $W$ was observed. In this paper, we build upon and extend the recent results of (Xu & Raginsky, 2020) to analyze the MER in Bayesian learning and derive information-theoretic bounds on it. We formulate the problem as a (constrained) rate-distortion optimization and show how the solution can be bounded above and below by two other rate-distortion functions that are easier to study. The lower bound represents the minimum possible excess risk achievable by \emph{any} process using $R$ bits of information from the parameter $W$. For the upper bound, the optimization is further constrained to use $R$ bits from the training set, a setting which relates MER to information-theoretic bounds on the generalization gap in frequentist learning. We derive information-theoretic bounds on the difference between these upper and lower bounds and show that they can provide order-wise tight rates for MER under certain conditions. This analysis gives more insight into the information-theoretic nature of Bayesian learning as well as providing novel bounds.
APA
Hafez-Kolahi, H., Moniri, B., Kasaei, S. & Baghshah, M.S.. (2021). Rate-Distortion Analysis of Minimum Excess Risk in Bayesian Learning. Proceedings of the 38th International Conference on Machine Learning, in Proceedings of Machine Learning Research 139:3998-4007 Available from https://proceedings.mlr.press/v139/hafez-kolahi21a.html.

Related Material