Probabilistic Non-Linear Principal Component Analysis with Gaussian Process Latent Variables

It is known that Principal Component Analysis has an underlying probabilistic representation based on a latent variable model. Principal component analysis (PCA) is recovered when the latent variables