Accuracy and macro F-Score (with 95% confidence intervals) of our different methods to capture case-level context on six different classification tasks using the HiSAN as the baseline.
收藏数据链接：
官方服务：
资源简介：
The top row is our baseline without any report level context, the middle group shows results of methods than can access both future and previous reports in a sequence, and the bottom group show results of methods that can only access previous reports in a sequence.
创建时间：
2020-05-12
