Multivariate Regression Estimation Based on the Maximum Likelihood Principle.
TEXAS A AND M UNIV COLLEGE STATION
Pagination or Media Count:
To estimate the population mean, Y bar, of a variate y a simple random sample of observations on y and several concomitant variables with known population means is considered. An approximation to the maximum likelihood estimator of Y bar is obtained and it is seen to be similar to the customary regression estimator. The result is a multivariate generalization of an earlier result due to Hartley and Rao 1968. Author
- Statistics and Probability