Model Discrimination for Nonlinear Regression Models

Borowiak (math, U. of Akron) discusses model discrimination based upon incorrect selection probability, presents diagnostic statistics and formal hypothesis test procedures to assess a model's fit and stability, explains the use of computer computations such as the jackknife and bootstrap, and demonstrates model discrimination techniques through examples taken from real experimental settings. Offering practical guidance that experimenters and students can apply to their own research problems, this definitive reference text is essential reading for statisticians, mathematicians, data analysts, forecasters, applied engineers and advanced undergraduate and graduate students in the above disciplines.