Overlapping Mixtures of Gaussian Processes for the Data Association Problem
Miguel Lazaro-Gredilla, Steven Van Vaerenbergh and Neil Lawrence
Universidad Carlos III de Madrid.
In this work we introduce a mixture of GPs to address the data association problem, i.e. to label a group of observations according to the sources that generated them. Unlike several previously proposed GP mixtures, the novel mixture has the distinct characteristic of using no gating function to determine the association of samples and mixture components. Instead, all the GPs in the mixture are global and samples are clustered following "trajectories" across input space. We use a variational Bayesian algorithm to efficiently recover sample labels and use a KL-corrected bound to learn the hyperparameters. We show how multiobject tracking problems may be disambiguated and explore the characteristics of the model in more traditional regression settings.