Saved in:
Bibliographic Details
Main Authors: Bakker, Hua Chang, Gupta, Shashank, Oosterhuis, Harrie
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2409.09819
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916436449951744
author Bakker, Hua Chang
Gupta, Shashank
Oosterhuis, Harrie
author_facet Bakker, Hua Chang
Gupta, Shashank
Oosterhuis, Harrie
contents Variance regularized counterfactual risk minimization (VRCRM) has been proposed as an alternative off-policy learning (OPL) method. VRCRM method uses a lower-bound on the $f$-divergence between the logging policy and the target policy as regularization during learning and was shown to improve performance over existing OPL alternatives on multi-label classification tasks. In this work, we revisit the original experimental setting of VRCRM and propose to minimize the $f$-divergence directly, instead of optimizing for the lower bound using a $f$-GAN approach. Surprisingly, we were unable to reproduce the results reported in the original setting. In response, we propose a novel simpler alternative to f-divergence optimization by minimizing a direct approximation of f-divergence directly, instead of a $f$-GAN based lower bound. Experiments showed that minimizing the divergence using $f$-GANs did not work as expected, whereas our proposed novel simpler alternative works better empirically.
format Preprint
id arxiv_https___arxiv_org_abs_2409_09819
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle A Simpler Alternative to Variational Regularized Counterfactual Risk Minimization
Bakker, Hua Chang
Gupta, Shashank
Oosterhuis, Harrie
Machine Learning
Variance regularized counterfactual risk minimization (VRCRM) has been proposed as an alternative off-policy learning (OPL) method. VRCRM method uses a lower-bound on the $f$-divergence between the logging policy and the target policy as regularization during learning and was shown to improve performance over existing OPL alternatives on multi-label classification tasks. In this work, we revisit the original experimental setting of VRCRM and propose to minimize the $f$-divergence directly, instead of optimizing for the lower bound using a $f$-GAN approach. Surprisingly, we were unable to reproduce the results reported in the original setting. In response, we propose a novel simpler alternative to f-divergence optimization by minimizing a direct approximation of f-divergence directly, instead of a $f$-GAN based lower bound. Experiments showed that minimizing the divergence using $f$-GANs did not work as expected, whereas our proposed novel simpler alternative works better empirically.
title A Simpler Alternative to Variational Regularized Counterfactual Risk Minimization
topic Machine Learning
url https://arxiv.org/abs/2409.09819