(A) task-oriented service personalization scheme for smart environments based on reinforcement learning

Abstract

In the domain of the Internet of Things (IoT), devices are made to understand and react to the current situation. These devices form smart environments, which are tasked with providing support to users. As users want to be provided with personalized support, these environments need to be able to learn user preferences, such as what temperature the room should be or which lights should be turned on. In this paper, we propose a novel system that separates the tasks of learning a user's preferences and realizing them within the environment. The system is able to capture user preferences by reinterpreting the problem as an optimization problem and applying inverse reinforcement learning to it. The system is shown to be able to accurately extract preferences related to each task given a small number of user demonstrations. These preferences are then realized by actuates devices running reinforcement learning-based agents to provide an environment consistent with the learnt preferences, even in situations not included in any user demonstration.