Timezone: »

 
Poster
Local Metric Learning for Off-Policy Evaluation in Contextual Bandits with Continuous Actions
Haanvid Lee · Jongmin Lee · Yunseon Choi · Wonseok Jeon · Byung-Jun Lee · Yung-Kyun Noh · Kee-Eung Kim

Wed Nov 30 02:00 PM -- 04:00 PM (PST) @ Hall J #406
in Poster Session 4 »

We consider local kernel metric learning for off-policy evaluation (OPE) of deterministic policies in contextual bandits with continuous action spaces. Our work is motivated by practical scenarios where the target policy needs to be deterministic due to domain requirements, such as prescription of treatment dosage and duration in medicine. Although importance sampling (IS) provides a basic principle for OPE, it is ill-posed for the deterministic target policy with continuous actions. Our main idea is to relax the target policy and pose the problem as kernel-based estimation, where we learn the kernel metric in order to minimize the overall mean squared error (MSE). We present an analytic solution for the optimal metric, based on the analysis of bias and variance. Whereas prior work has been limited to scalar action spaces or kernel bandwidth selection, our work takes a step further being capable of vector action spaces and metric optimization. We show that our estimator is consistent, and significantly reduces the MSE compared to baseline OPE methods through experiments on various domains.

Keywords: contextual bandit · Local Kernel Metric Learning · Continuous Actions · Importance Sampling · Off-Policy Evaluation
[ Paper]  [ Slides]  [ Poster]  [ OpenReview] 

Author Information

Haanvid Lee (KAIST)
Jongmin Lee (UC Berkeley)
Yunseon Choi (Korea Advanced Institute of Science & Technology)
Wonseok Jeon (Qualcomm AI Research)
Byung-Jun Lee (KAIST)
Yung-Kyun Noh (Hanyang University / Korea Institute for Advanced Study)
Kee-Eung Kim (KAIST)

More from the Same Authors