search

JuliaStats / MultivariateStats.jl

A Julia package for multivariate statistics and data analysis (e.g. dimension reduction)
Other
377 stars 86 forks source link

Support missing data in PPCA #138

Open oxinabox opened 3 years ago

oxinabox commented 3 years ago

According to the PPCA paper

Tipping, Michael E., and Christopher M. Bishop. "Probabilistic principal component analysis." Journal of the Royal Statistical Society: Series B (Statistical Methodology) 61.3 (1999): 611-622.  one of the key advantages of PPCA over PCA is that it can naturally handle missing data.    The method is in section 4.1 of the paper; though it's not a great decription.  It would be cool to have.

wildart commented 3 years ago

It's more about the estimation of Gaussian model properties from the missing data, see section 11.2 in "Statistical Analysis with Missing Data",Little and Rubin, 2020.