An Experimental Analysis for Detecting Wi-Fi Network Associations Using Multi-label Learning

Buildings play an important role in our lives, although they require great amounts of energy and resources to operate. The creation of cheaper and more efficient smart building management systems to control building infrastructure as lighting and HVAC (Heating,Ventilating and Air Conditioning) is required. Those systems should correctly identify users inside building areas for only providing the necessary resources. Different sensor networks can be used to correctly detect people inside buildings areas. Wi-Fi networks have attracted a lot of attention due to its large deployment. In this work, we evaluate the performance of Multi-label machine learning methods to build classification models capable of providing wireless network access points (AP) occupancy predictions throughout the day. The APs association history used in our experiment was built using data from SCIFI network of Fluminense Federal University for a period of 6 months. Our results show that the best overall achieved accuracy was 87.04% for predicting access point occupancy.