Human-Machine Co-Learning
Anticipating, Identifying and Sharing Emergent Collaboration Patterns
E.M. van Zoelen (TU Delft - Electrical Engineering, Mathematics and Computer Science)
M.A. Neerincx – Promotor (TU Delft - Electrical Engineering, Mathematics and Computer Science)
D.A. Abbink – Promotor (TU Delft - Mechanical Engineering)
More Info
expand_more
Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.
Abstract
Intelligent machines (in the form of physically embodied robots or virtual agents) are increasingly able to perform tasks in collaboration with humans. However, learning to become a good team takes time, especially when dynamic tasks require the team to constantly adapt to new situations. Over time, both human and machine need to not only learn how to execute the task, but also how their team partner behaves in the task, as well as how to improve their collaboration over time by attuning their behavior to each other. Existing research on human-machine collaboration often does not sufficiently address adaptation and learning. Work that does study adaptation and learning tends to focus on either machine learning and adaptation or human learning and adaptation, thereby not addressing the interaction of these learning processes that would be present in a co-learning situation...