Simultaneous Estimation by Partial Totals for Compartmental Models
John J. Beauchamp, Richard G. Cornell · Journal of the American Statistical Association · 1968
An estimation procedure is presented that may be used to simultaneously estimate the parameters in a multiple equation regression model.The regression models considered are shown to arise from radioactive tracer experiments using compartmental models.models the regression equations are shown to be linear combinations of the same exponential parameters, and the number of independent regression equations is also shown to be the same as the number of exponential parameters. For suchThe estimation procedure is developed under the assumption of equally spaced values of the independent variable.The first step of the procedure involves the simultaneous estimation of the exponential parameters by a generalized patial totals approach, and the second step involves the simultaneous estimation of the linear parameters using the estimates of the nonlinear parameters.presented that make the estimation procedure more versatile, and asymptotic properties are also investigated.with generated data.