Title: A clustering ensemble framework for learning mixtures of Gaussian processes Authors:  Mimi Zhang - Trinity College Dublin (Ireland) [presenting]
Abstract: A clustering ensemble framework is developed that can efficiently identify the latent cluster labels of functional data from a Gaussian process mixture. This approach exploits the independence and Gaussianity of the coefficients in the Karhunen-Loeve expansion of a Gaussian random function. In particular, each base clustering in the ensemble is obtained by fitting a univariate Gaussian mixture model to the projection coefficients of the functional data onto one basis function, and the different basis functions are orthonormal. The computational complexity for identifying the cluster labels is much lower than state-of-the-art methods, and theoretical guarantees on the quality of the consensus clustering are provided. Extensive experimental studies on both synthetic and real datasets demonstrate that our method outperforms existing methods.