The original dataset of "Prediction of density and viscosity and identification of key features of biofuel-alcohol blends based on explainable machine learning models"

The density and viscosity experimental methods used were vibration tube method and capillary tube method. After using the Z-score method to clean the experimental data for outliers, the dataset is divided into a training set, a testing set, and a validation set.

The validation set is used to monitor early stopping conditions, while the testing set does not participate in model training and validation, and is used as new unknown data for final performance evaluation.