Year: 2019
Type: conference-paper
Abstract: This paper proposes to incorporate uncertainty and performance into AI-advised human decision making to boost user trust in AI.Motivated by the trial-to-trial approach in the adjustment of decision boundaries by humans in decision making, this paper presents a novel framework of human-AI teaming for trustworthy AI under uncertainty.The framework employs the trial-to-trial approach to simulate the trial to trial process for adjusting decision boundaries and provide an estimation of uncertainty in AI.An Uncertainty-Performance Interface (UPI) in the framework is then proposed to allow users access the quality and performance of machine learning models in AI at the same time.The proposed framework allows human and AI collaborate in a teaming environment for trustworthy decisions under uncertainty.
Authors: Zhou, J, Chen, F
Institution: University of Technology Sydney
Language: English
Subfield: Human Factors and Ergonomics
Field: Social Sciences
Domain: Social Sciences
Open Access status: green
Award: G1380602548
OpenAlex ID: W7043719198