Handling missing data on asymmetric distribution

The problem of imputation of missing observations emerges in many areas. Datausually contained missing observations due to many factors, such as machine failuresand human error. Incomplete dataset usually causes bias due to differences betweenobserved and unobserved data. This paper proposed Neyman allocation method toestimate asymmetric winsorizing mean for handling missing observations when thedata follow the exponential distribution. Different values of the exponentialdistribution parameters were used to illustrate. A set of data from exponentialdistribution were generated to compare the performance of the proposed methodssuch as regression trend, average of the whole data, naive forecast and average boundof the holes and the proposed Neyman allocation method. The goodness-of-fitcriterions used were the mean absolute error (MAE) and the mean squared error(MSE). It was found that the proposed method gave the best fit in the sense of havingsmaller error, in particular for a large percentage of missing observations.