Machine learning (ML) models are increasingly being used in application domains that often involve working together with human experts. In this context, it can be advantageous to defer certain instances to a single human expert when they are difficult to predict for the ML model. While previous work has focused on scenarios with one distinct human expert, in many real-world situations several human experts with varying capabilities may be available. In this work, we propose an approach that trains a classification model to complement the capabilities of multiple human experts. By jointly training the classifier together with an allocation system, the classifier learns to accurately predict those instances that are difficult for the human experts, while the allocation system learns to pass each instance to the most suitable team member -- either the classifier or one of the human experts. We evaluate our proposed approach in multiple experiments on public datasets with "synthetic" experts and a real-world medical dataset annotated by multiple radiologists. Our approach outperforms prior work and is more accurate than the best human expert or a classifier. Furthermore, it is flexibly adaptable to teams of varying sizes and different levels of expert diversity.
Exact Algorithms and Lower Bounds for Forming Coalitions of Constrained Maximum Size
Department of Theoretical Computer Science, Faculty of Information Technology(理论计算机科学系);Université Clermont Auvergne, CNRS, Clermont Auvergne INP, Mines Saint-Étienne, LIMOS(克莱蒙特-奥弗涅大学);G-SCOP, Grenoble-INP(格勒诺布尔INP);Computer Science Institute, Faculty of Mathematics and Physics(数学与物理系计算机科学研究所)
Advancing Human-AI Complementarity: The Impact of User Expertise and Algorithmic Tuning on Joint Decision Making
Microsoft Research(微软研究院);University of Virginia(弗吉尼亚大学);University of Washington(华盛顿大学);University of Michigan(密歇根大学);Human Computation Institute(人类计算研究所);LMU Munich(慕尼黑大学);Western Washington University(西华盛顿大学)