A Bayesian Approach to Nonlinear Latent Variable Models using the Gibbs Sampler and the Metropolis-Hastings Algorithm | Synapse