Imputation method in the presence of censored data. The main message of the imputation method is that we should account for the variability of the censored part of the data by mimicking the complete data. That is, we impute the incomplete part with a conditional random sample rather than the conditional expectation or certain constants. Simulation results suggest that the imputation method reduces the possible biases and has similar standard errors than those from complete data. … Censored Time Series Analysis google

Advertisements