In section 4.2 of Bishop and Tipping's "Probabilistic Principal Component Analysis" (Microsoft Research), we are shown three different plots that are said to have something to do with a three-component mixture model of probabilistic PCA projections. I find it hard to understand exactly what is meant by this.

It seems like the same data set was used in each plot. So if the data is the same, what is done differently in each plot to give varying appearances? The explanation is very limited. Does each plot correspond to its own PCA projection? In that case, does the difference in appearance stem from the use of different principal components (eigenvectors in the basic PCA formulation) in each plot? If so, how were these selected?


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.