Package R for "Stochastic Approximation EM for Logistic Regression with Missing Values (W. Jiang, J. Josse, M. Lavielle, 2018)"
misaem
is a method to apply statistical inference for logistic regression model with missing data. This method is based on likelihood, including
- A stochastic approximation version of EM algorithm based on Metropolis-Hasting sampling, to estimate the parameters of logistic regression;
- Estimation of parameters' variance based one Louis formula;
- Model selection procedure based on AIC or BIC.