In this thesis the five equation system of growth and yield equations originally developed by Clutter (1963) is examined. The system is redeveloped algebraically to form a truly algebraically compatible system.
Three methods of estimating the coefficients were examined. In the first method, three of the equations were fitted independently using ordinary least squares; these coefficient estimates were carried through to the other equations. No consideration was given to the relationships that must exist between the equation coefficients in order for the system to be numerically consistent. In the second method the system is first developed algebraically, before any of the coefficients are estimated, resulting in a slightly different system which is truly algebraically compatible. The coefficients were estimated by fitting two of the equations, and using these estimates throughout the rest of the system. The resulting system is both numerically consistent and algebraically compatible. In the final method the relationships between the coefficients that must hold for the system to be compatible were incorporated in the coefficient estimation procedure. Seemingly unrelated regression techniques were used to estimate the coefficients.
The three methods resulted in coefficient estimates that were similar, with seemingly unrelated regression producing the most efficient estimators. Prediction ability of the three methods on independent data show no method as being superior, although the seemingly unrelated regression procedure was able to reduce the total system error best. / M.S.
Identifer | oai:union.ndltd.org:VTETD/oai:vtechworks.lib.vt.edu:10919/94460 |
Date | January 1986 |
Creators | Hans, Richard P. |
Contributors | Forestry |
Publisher | Virginia Polytechnic Institute and State University |
Source Sets | Virginia Tech Theses and Dissertation |
Language | en_US |
Detected Language | English |
Type | Thesis, Text |
Format | vii, 46 leaves, application/pdf, application/pdf |
Rights | In Copyright, http://rightsstatements.org/vocab/InC/1.0/ |
Relation | OCLC# 15255701 |
Page generated in 0.0021 seconds