Dalin Zeng, Youquan Xu, Xiancu
Abstract
Assessing health risk factors often involves observational data that with heterogeneous patient background that information is limited to researchers. Thus forms a curvature, sometimes in complicate shape, on the continuous predictive variables in fitting models. Fractional Polynomials (FP) transformation provide a flexiblemodel fitting framework account such curvature. In this paper we assess the utilization of FP model fitting framework in health risk factor assessment.