Comparison of Data Grouping Strategies on Prediction Accuracy of Tree-Stem Taper for Six Common Species in the Southeastern US

Clustering data into similar characteristic groups is a commonly-used strategy in model development. However, the impact of grouping strategies on modeling stem taper has not been well quantified. The objective this study was to compare prediction accuracy different strategies. Specifically, population-level compared models fitted with grouped based taxonomic rank, tree form and size. A total 3678 trees were used analyses, which included six common species upland hardwood forests southeastern U.S. Results showed that overall predictions are more accurate when building at species, group or division level rather than population level. considerably improved between species-specific functions species-related for four examined. Grouping by rank provided reliable height-to-diameter ratio (H–D ratio) diameter breast height (DBH). form/size-related methods (i.e., H–D DBH) generally did improve precision model. In study, effect sample size fitting minimal accuracy. methodology presented provides mixed-species data, will be practical importance needed developing models.