Enhancing Access to Machine-Learning Models. We packaged our  electronic  classifiers  and  made  them  publically  available.  They are  easily  accessible  via  an  interactive Jupyter notebookhosted  by Binder.  Anyone  can  upload  a  structure  file  in  CIF  format  and  make their own prediction using the interactive Jupyter notebook. Since this notebook  is  hosted  in  a  Docker  containerized  environment,  any person  interested  in  making  a  classification  can  execute  the  script immediately in their web browser without installing any dependencies. This  aspect  greatly  improves  the  usability  of  the  code  and  broadens electronic    materials    research    participation,    especially    to    non-computational  researchers  and  non-domain  experts.  The  complete workflow   behind   the   ML   models   is   described   in   the   project’s GitHub pagewith   some   sub-functions   also   demonstrated   in   an interactive Jupyter notebook.