In this repository, we keep some of the scripts and data used during our investigation about glass property prediction and design of new glasses (see articles below). We have developed machine learning models capable of predicting the most common properties of glass with high performance. In addition, we created several scripts to optimize the models generated by the machine learning algorithms and to understand the knowledge acquired by these models (explainability).
Unfortunately, we failed to put explicit comments in some codes while they were being developed. Sorry, this can cause you to waste additional time trying to understand the codes. If you want to do a full reading, I recommend starting with Makefile. There we describe how each script was executed on our servers.
Below are some articles where we describe the results obtained. Please note that we make versions available on Arxiv, in case you do not have access to these journals.
Alcobaça, E., Mastelini, S. M., Botari, T., Pimentel, B. A., Cassar, D. R., de Leon Ferreira, A. C. P., & Zanotto, E. D. (2020). Explainable machine learning algorithms for predicting glass transition temperatures. Acta Materialia, 188, 92-100. (journal)(open-access)(sup-material)
Cassar, D. R., Mastelini, S. M., Botari, T., Alcobaça, E., de Carvalho, A. C., & Zanotto, E. D. (2021). Predicting and interpreting oxide glass properties by machine learning using large datasets. Ceramics International. (journal)(open-access)
Mastelini, S. M., Cassar, D. R., Alcobaça, E., Botari, T., de Carvalho, A. C., & Zanotto, E. D. (2021). Machine learning unveils composition-property relationships in chalcogenide glasses. arXiv preprint arXiv:2106.07749. (journal)(open-access)