FT-IR data set for hierarchical cluster-based deep learning

FT-IR of raw materials described in Kristoffersen et al. (2019) doi:10.1016/j.talanta.2019.06.084. Predicts average molecular weight after hydrolysis. The data is pre-processed using Savitzky-Golay 2nd derivative smoothing (with window width 11 pt and 3rd order polynomial smoothing) followed by extended multiplicative signal correction (EMSC).本数据集为Kristoffersen等人（2019）在DOI:10.1016/j.talanta.2019.06.084的研究中描述的原料傅里叶变换红外光谱（FT-IR）数据，可用于预测水解后的平均分子量。所有数据均经过预处理：首先采用萨维茨基-戈莱（Savitzky-Golay）二阶导数平滑法（窗口宽度为11点，采用三阶多项式平滑），随后执行扩展多元信号校正（EMSC）。