Sensor Selection in Smart Homes

There has been a lot of work focussing on activity recognition in smart homes, with the aim of the home being to monitor the activities of the inhabitant and identify deviations from the norm. For a smart home to support its inhabitants, the recognition sys- tem needs to accurately learn from the observations acquired through sensors, which are installed in the home. Given a predefined set of the inhabitant's daily activities, the question is which sensors are important to accurately recognise these activities. This paper addresses the sensor selection problem through a filter-based approach, which is based on information gain. We evaluate the effectiveness of the proposed method on two publicly available smart home datasets.