We consider active learning with logged data, where labeled examples are drawn conditioned on a predetermined logging policy, and the goal is to learn a classifier on the entire population, not just conditioned on the logging policy. Prior work addresses this problem either when only logged data is available, or purely in a controlled random experimentation setting where the logged data is ignored. In this work, we combine both approaches to provide an algorithm that uses logged data to bootstrap and inform experimentation, thus achieving the best of both worlds. Our work is inspired by a connection between controlled random experimentation and active learning, and modifies existing disagreement-based active learning algorithms to exploit logged data.
Author Information
Songbai Yan (University of California San Diego)
Kamalika Chaudhuri (University of California at San Diego)
Tara Javidi (University of California San Diego)
Related Events (a corresponding poster, oral, or spotlight)
-
2018 Oral: Active Learning with Logged Data »
Fri Jul 13th 05:50 -- 06:00 PM Room A6
More from the Same Authors
-
2019 Talk: Opening Remarks »
Kamalika Chaudhuri · Ruslan Salakhutdinov -
2019 Tutorial: Active Hypothesis Testing: An Information Theoretic (re)View »
Tara Javidi -
2018 Poster: Analyzing the Robustness of Nearest Neighbors to Adversarial Examples »
Yizhen Wang · Somesh Jha · Kamalika Chaudhuri -
2018 Oral: Analyzing the Robustness of Nearest Neighbors to Adversarial Examples »
Yizhen Wang · Somesh Jha · Kamalika Chaudhuri -
2017 Workshop: Picky Learners: Choosing Alternative Ways to Process Data. »
Corinna Cortes · Kamalika Chaudhuri · Giulia DeSalvo · Ningshan Zhang · Chicheng Zhang -
2017 Poster: Active Heteroscedastic Regression »
Kamalika Chaudhuri · Prateek Jain · Nagarajan Natarajan -
2017 Talk: Active Heteroscedastic Regression »
Kamalika Chaudhuri · Prateek Jain · Nagarajan Natarajan
