The Effect of Co-adaptive Learning & Feedback in Interactive Machine Learning ZbyszynskiMichael Di DonatoBalandino TanakaAtau 2020 In this paper, we consider the effect of co-adaptive learning on the training and evaluation of real-time, interactive machine learning systems, referring to specific examples in our work on action-perception loops, feedback for virtual tasks, and training of regression and temporal models. Through these studies we have encountered challenges when designing and assessing expressive, multimodal interactive systems. We discuss those challenges to machine learning and human-computer interaction, proposing future directions and research.