B1408
Title: Functional survival analysis with convex clustering
Authors: Yuko Araki - Shizuoka University (Japan) [presenting]
Abstract: The survival analysis which contains several classes of functional process, measured at baseline, is investigated. In our method, first, we identify the class of individual trajectory by the proposed functional convex clustering method, and as a second stage, class information is used in the Cox proportional hazards regression models to assess the risk of mortality during the follow-up period. We assess the performance of the proposed model in a simulation study. For illustration, the proposed method was applied to the cohort study in Japan.