TY  - BOOK
AU  - Alaa Sayed Shehata
AU  - Ahmed Amin Elsheikh , 
AU  - Naglaa Abdelmoneim Morad , 
TI  - Estimation of some distribution parameters with incomplete data / 
PY  - 2014///
CY  - Cairo : 
PB  - Alaa Sayed Shehata , 
KW  - Distribution parameters
KW  - Estimation
KW  - Incomplete data
N1  - Thesis (M.Sc.) - Cairo University - Institute of Statistical Studies and Research  - Department of Statistics and Econometrics; Issued also as CD
N2  - There are small differences between mean square error in case of the three methods (maximum likelihood method, listwise deletion method and mean imputation).  The difficulty of obtaining the values of estimators (k_1,k_2,k) using maximum likelihood method, in this method, it is replaced missing data by zero and because these estimators consist of log (variable), log (zero) is not defined. But it is possible to find a solution to this problem by removing the row which consists of missing data i.e. replacing maximum likelihood method by listwise deletion method.  If you are interested in parameter k, It is better to use mean imputation method. If you are interested in parameter k, It is better to use listwise deletion method.  When percentage of missing data increases, for parameter k, It is better to use mean imputation or maximum likelihood method. For parameter k, It is better to use listwise deletion method. When sample size increases, for parameter k, It is better to use mean imputation method. For parameter k, It is better to use listwise deletion method. For Geometric distribution: There are small differences between mean square error in case of three methods (maximum likelihood method, listwise deletion method and mean imputation). It is better to use mean imputation method i.e the mean square error of parameters is the smallest using mean imputation method
UR  - http://172.23.153.220/th.pdf
ER  -