MDIr

Overfitted Bayesian mixture models and Multiple Dataset Integration. Kernels allowed are

• G: Gaussian (independent measurements/diagonal covariance matrix).
• MVN: Multivariate normal (full covariance matrix).
• C: Categorical.
• GP: Gaussian process (squared exponential covariance kernel).
• TAGM: t-augmented Gaussian (MVN with a global MVT component for outliers).
• TAGPM: t-augmented Gaussian Process (GP with a global MVT component for outliers).