Comparison methods of estimating missing data in real data time series

Missing data are encountered in many researches and they are also found in well-conducted and controlled studies. Missing data can reduce the statistical strength of a study and may produce biased estimates, leading to invalid conclusions. This study is focused on the problems and types of missing data, together with the techniques for their approach. The mechanisms by which the missing data are obtained and the methods to study these data are illustrated. We have dealt with the multiple imputations as a very efficient method of imputing the missing data and applying these methods in some simulation cases and in real data time series. We have also prepared and adapted the scripts in the programming language R to conduct the simulations. The proposed mice and Amelia packages for imputing the missing values provide fairly good approximations even in the case of real data.