Estimating Missing Values in Multivariate-Time-Series Clinical Data using Gradient Boosting Tree on Temporal and Cross-variable Features
Xu Xiao, Junmei Wang, Xian Xu, Yuyao Sun, Quanhe Chen, Xiang Li, Guotong Xie · 2019
Estimation of missing values in multivariate-time-series clinical data is of great importance for data analysis. It is common to use mean value and forward/backward value of a time-series to filling in the missing parts. However, the simply statistical values can hardly capture the rich information in multi-time-series for estimating miss values.