Effects of case removal in prognostic models

Constructing and updating prognostic models that learn from training cases is a time-consuming task. The more compact, and yet informative, the traini ng sets a re, the faster one can build and properly evaluate such models. W e have compared different regression diagnostic methods for selection and r emoval of training cases in prognostic models. Univariate determinations we re performed using classical regression diagnostic statistics. Multivariate determinations were performed using (1) a sequential "backward" selection of cases, and (2) a non-sequential genetic algorithm. The genetic algorithm produced final models that kept few cases and retained predictive capabili ty. A genetic algorithm approach to case selection may be better suited for guiding removal of cases in training sets than a univariate or a sequentia l multivariate approach, possibly because of its ability to defect sets of cases that are influential en bloc but may not be sufficiently influential when considered in isolation.