A novel application of the expectation maximization (EM) algorithm is proposed for modeling rightcensored multiple regression. Parameter estimates, variability assessment, and model selection are summarized in a multiple ...