ClarusC64/clinical-renal-buffer-exhaustion-v0.1

该数据集用于评估模型是否能检测由肾脏缓冲衰竭引起的不稳定性。每个数据行代表一个简化的肾脏调节场景，包含三个时间点的观察数据。任务目标是判断肾脏缓冲是保持稳定还是趋向于不稳定。数据集的核心稳定性理念涉及肾脏通过过滤、排泄和液体调节来稳定代谢和电解质平衡。不稳定的情况包括尿量下降、肌酐上升、钠调节不稳定等。数据集的结构包括尿量轨迹、肌酐代理轨迹、钠代理轨迹等变量，并包含干扰变量。预测目标为标签1表示肾脏缓冲衰竭不稳定，标签0表示稳定的肾脏调节。

This dataset evaluates whether models can detect instability caused by renal buffering failure. Each row represents a simplified renal regulation scenario observed across three time points. The task is to determine whether renal buffering remains stable or is moving toward renal instability. The core stability idea involves the kidney stabilizing metabolic and electrolyte balance through filtration, excretion, and fluid regulation. Instability emerges when urine output declines, creatinine rises, sodium regulation destabilizes, etc. The dataset structure includes urine output trajectory, creatinine proxy trajectory, sodium proxy trajectory, and decoy variables. The prediction target is label = 1 for renal buffer exhaustion instability and label = 0 for stable renal regulation.