Modeling missing covariate data and temporal features of time-dependent covariates in tree-structured survival analysis

摘要: Tree-structured survival analysis (TSSA) is used to recursively detect covariate values that best divide the sample into subsequent subsets with respect a time event outcome. The result set of empirical classification groups, each which identifies individuals more homogeneous risk than original sample. We propose methods for managing missing data and also incorporating temporal features repeatedly measured covariates TSSA. First, data, we an algorithm uses stochastic process add draws existing single tree-structured imputation method. Secondly, incorporate covariates, two different methods: (1) use two-stage random effects polynomial model estimate be as TSSA predictor variables, (2) other types functions time-dependent methodology. conduct simulation studies assess accuracy predictive abilities our proposed Our methodology has particular public health importance because create, interpret algorithms can in clinical setting predict response treatment late-life depression.