BNP with Turing.jl

We recently extended Turing.jl to be used for non-parametric probabilistic modelling. This blog post briefly summarizes our approach. Probabilistic modelling is a core component of a scientists’ toolbox for incorporating uncertainties about the model parameters and noise in the data. Statistical models with a Bayesian nonparametric (BNP) component are difficult to handle due to the infinite dimensionality… Continue reading BNP with Turing.jl