1 Department of Applied Mathematics and Computer Science, Technical University of Denmark2 Cognitive Systems, Department of Applied Mathematics and Computer Science, Technical University of Denmark3 Copenhagen Center for Health Technology, Center, Technical University of Denmark
We propose the following generalization of the Variational Garrote for sequential EEG imaging: A Markov prior to promote sparse, but temporally smooth source dynamics. We derive a set of modied Variational Garrote updates and analyze the role of the prior's hyperparameters. An experimental evaluation is given in simulated data and in a benchmark EEG data set.
Proceedings of the 3rd Nips Workshop on Machine Learning and Interpretation in Neuroimaging 2013, 2013
Main Research Area:
27th Annual Conference on Neural Information Processing Systems (NIPS 2013)Neural Information Processing Systems