Skip to main content
Fig. 3 | BMC Medical Informatics and Decision Making

Fig. 3

From: Sparse multi-output Gaussian processes for online medical time series prediction

Fig. 3

The trained kernel for one patient jointly modeling PT and INR. For both the (a) SM-LMC kernel and (b) sparse SM-LMC kernel, the Aq matrices (upper row), Bq matrices (middle row), and the basis kernel Îşq (bottom row) are illustrated. The zero elements are colored in light grey. Here l.c denotes length scale for each basis kernel, and per. denotes period. The length-of-stay for this patient was over 90 days

Back to article page