An Adaptive Approach for Hazard Regression Modeling

Regression models for survival time data involve estimation of the hazard rate as a function predictor variables and associated slope parameters. An adaptive approach is formulated such regression modeling. The modeled using fractional polynomials, that is, linear combinations products power transforms together with other available predictors. These polynomial are restricted to generating positive-valued rates decreasing times. Exponentially distributed times special case. Parameters estimated maximum likelihood allowing right censored Models evaluated compared cross-validation (LCV) scores. LCV scores tolerance parameters used control an search through alternative identify effective underlying data. methods demonstrated two different sets including lung cancer patients multiple myeloma patients. For data, depends distinctly on time. However, controlling cell type provides distinct improvement while only no longer Furthermore, Cox unable effect. also Moreover, consideration hemoglobin at diagnosis improvement, still time, moderates effect rate. results indicate modeling can provide unique insights into