Quantile Regression Approach to Estimating Percentile Growth Model

Chengcai Ni, Chris J. Cieszewski, Bruce E. Borders

D. B. Warnell School of Forest Resources

University of Georgia

Athens, GA, 30602

Abstract:
The percentile growth model plays an important role in modeling stand diameter distribution changes through time. We proposed the quantile regression method to estimate coefficients of percentile growth models due to its’ simplicity and efficiency in terms of the first and second order. Statistical simulation showed the new estimation method converges to the real value of parameters faster than ordinary least squares with smaller variance in cases where conditional diameter distribution is heteroscedastic.

Author Keywords:

quantile regression, diameter distribution, percentile growth model, stand table, statistical simulation, forest.

KeyWords Plus:

Addresses:

Publisher:

IDS Number:

ISSN: