Hi guys,

I would like to get the idea on how to implement the conditional mean imputation in the Finite Gaussian Mixture model, say 2 Gaussian components .This is because, if i bring the two mean and sigma values in the e-step to fill-in the missing data, i'm a bit confuse on how to impute missing data that involves with two components. This is the snippet code of sample conditional means imputation:

imp1

More Aisyah mat jasin's questions See All
Similar questions and discussions