Provided by: The Pennsylvania State University
Date Added: Oct 2011
In this paper, the authors present the in-context application for Smartphones, which combines signal processing, active learning and reinforcement learning to autonomously create a personalized model of interruptibility for incoming phone calls. They empirically evaluate the system, and show that they can obtain an average of 96.12% classification accuracy when predicting interruptibility after a week of training. In contrast to previous paper, they leverage density-weighted uncertainty sampling combined with a reinforcement learning framework applied to passively collected data to achieve comparable or superior classification accuracy using many fewer queries issued to the user.