Open kjappelbaum opened 2 years ago
currently, the most common way to include the nets in ML for MOFs is some kind of categorical encoding (e.g. one-hot). However, we can also view (as done in moffragmentor) the net as StructureGraph and then simply compute features on it.
moffragmentor
StructureGraph
currently, the most common way to include the nets in ML for MOFs is some kind of categorical encoding (e.g. one-hot). However, we can also view (as done in
moffragmentor) the net asStructureGraphand then simply compute features on it.