This thesis proposes a sensor failure detection and isolation system for AAL environments equipped with event-driven, ambient, binary sensors. First, an extensive literature review of research works was conducted. Then, the feasibility of extracting sensor correlations using the association rule mining technique was studied. Finally, a sensor failure detection and isolation system was developed based on the extracted correlations. Guidelines for selecting the values of the parameters of the system were also presented. The proposed approach was evaluated on two publicly available datasets.
«
This thesis proposes a sensor failure detection and isolation system for AAL environments equipped with event-driven, ambient, binary sensors. First, an extensive literature review of research works was conducted. Then, the feasibility of extracting sensor correlations using the association rule mining technique was studied. Finally, a sensor failure detection and isolation system was developed based on the extracted correlations. Guidelines for selecting the values of the parameters of the syst...
»