An EM algorithm for GMM parameter estimation in the presence of censored and dropped data with potential application for indoor positioning

Abstract

In this paper, a specific type of incomplete data in Wi-Fi fingerprinting based indoor positioning systems (WF-IPS) is presented: censored and dropped mixture data. For fitting this type of data, a censored and dropped Gaussian Mixture Model (CD-GMM) was proposed. Further, an extended version of the Expectation–Maximization (EM) algorithm is developed for estimating parameters of this model. Simulation results show the advantage of our proposal compared to existing methods. Thus, this approach not only has potential for the WF-IPSs, but also for other applications. Keywords: Expectation–Maximization, Gaussian Mixture Model, Censored and dropped data, Indoor positioning, Fingerprinting