Title: Functional Cox model
Authors: Xingqiu Zhao - The Hong Kong Polytechnic University Shenzhen Research Institute (China)
Meiling Hao - The Hong Kong Polytechnic University (Hong Kong)
Kin Yat Liu - The Hong Kong Polytechnic University (Hong Kong) [presenting]
Abstract: Right-censored data in the presence of both functional and scalar covariates often occur in many biomedical studies. We apply a roughness regularization approach in making nonparametric inference for functional Cox models. In a reproducing kernel Hilbert space framework, we construct asymptotically confidence intervals for functional and scalar covariates and two statistical procedures for hypothesis testing with integro-differential equation techniques and functional Bahadur representation. The finite sample performance of the proposed inference procedures are evaluated through simulation studies. The proposed method is illustrated with a study of association between daily measures of the Intensive Care Unit (ICU) Sequential Organ Failure Assessment (SOFA) score and mortality.