Fig. 2From: Sparse multi-output Gaussian processes for online medical time series predictionIllustrations of the basis kernels and the features for kernel clustering. (a) An example of a discrete set of basis kernels with different ÎĽ and v within a 72-h window. (b) An example of the 72-dim temporal features (shown with red dots) taken from a kernel for GMM clusteringBack to article page