Application of Constrained Optimization Approach to Missing Data in Experimental Design

During the course of a research data obtained may be fully observed or particularly observed. If the data obtained from a research is partially observed, then a common problem in experiment has occurred. This problem is known as the missing observation. Missing observation infers that no data (value) is stored for the variable in the current observation. Missing data are recurring in all sorts of research irrespective of the field, science, medical, agricultural and social science and so on. Researchers are faced with the problem of partially observed data sets. There are several reasons why data may be missing. They may be missing due to failure to record, gross errors in recording, accident and death amongst others. Missing data are very sensitive issues and many analyses techniques cannot proceed with gaps in their data. These missing values must be estimated and replaced before the analysis can be completed.