Potential prediction bias in regression and discriminant analysis
- 1 December 1986
- journal article
- research article
- Published by Canadian Science Publishing in Canadian Journal of Forest Research
- Vol. 16 (6) , 1255-1257
- https://doi.org/10.1139/x86-222
Abstract
Prediction bias is the difference between a model's apparent and actual prediction errors. Prediction bias is likely to occur when a model contains many independent variables relative to sample size or when many different sets of independent variables are tested by a stepwise procedure. Examples of potential prediction bias are illustrated by comparing published models with models developed using random numbers. Model prediction bias can be estimated by using a resampling procedure called the bootstrap. The bootstrap procedure is illustrated with a simple example.Keywords
This publication has 2 references indexed in Scilit:
- Site Quality Evaluation for White Spruce Plantations Using Discriminant AnalysisSoil Science Society of America Journal, 1985
- Vegetational indicators as independent variables in forest growth prediction in West-Central Alberta, CanadaForest Ecology and Management, 1984