Training Performance, Dataset Classification and Prediction Outputs of Hybrid DL Model for EI-MS Prediction

This dataset contains the numerical data used to generate figures in the related article “Hybrid Deep Learning Model for EI-MS Spectra Prediction” (DOI: 10.3390/ijms27031588). It includes model training data and evaluation metrics (ten random seeds) for baseline and ablated model variants, as well as spectral outputs from the best-trained model for three molecules: acetylene, 4-chloro-3-nitrobenzophenone, and 1,2,3,4-tetraphenyl-1,4-butanedione. For 4-chloro-3-nitrobenzophenone, prediction results from two additional models (NEIMS and RASSP SubsetNet) are also provided. The dataset further contains classification results that assign training and test compounds to molecular classes and superclasses. A metadata file is included to provide details on the contents of each csv file.