Skip to yearly menu bar Skip to main content


Poster

SEL-BALD: Deep Bayesian Active Learning for Selective Labeling with Instance Rejection

Ruijiang Gao · Mingzhang Yin · Maytal Saar-Tsechansky

Poster Room - TBD
[ ]
Fri 13 Dec 11 a.m. PST — 2 p.m. PST

Abstract:

Machine learning systems are widely used in many high-stakes contexts in which experimental designs for assigning treatments are infeasible. When evaluating a decision instance is costly, such as investigating a fraud case, or evaluating a biopsy decision, a sample-efficient strategy is needed. However, while existing active learning methods assume humans will always label the instances selected by the machine learning model, in many critical applications, humans may decline to label instances selected by the machine learning model due to reasons such as regulation constraint, domain knowledge, or algorithmic aversion, thus not sample efficient. In this paper, we propose the Active Learning with Instance Rejection (ALIR) problem, which is a new active learning problem that considers the human discretion behavior for high-stakes decision making problems. We propose new active learning algorithms under deep Bayesian active learning for selective labeling (SEL-BALD) to address the ALIR problem. Our algorithms consider how to acquire information for both the machine learning model and the human discretion model. We conduct experiments on both synthetic and real-world datasets to demonstrate the effectiveness of our proposed algorithms.

Live content is unavailable. Log in and register to view live content