2017/11/29 by Eric Gossett, Cormac Toher, Gossett, Eric +19 · 1 citation
Materials Science · #Computational Physics (physics.comp-ph) #Electron and X-Ray Spectroscopy Techniques #FOS: Physical sciences #Machine Learning in Materials Science #Materials Science (cond-mat.mtrl-sci) #X-ray Diffraction in Crystallography
paper · pdf · doi:10.48550/arxiv.1711.10744
openalex publication_date 2017/11/29 · openalex created_date 2022/10/01 · openalex updated_date 2026/07/28
Machine learning approaches, enabled by the emergence of comprehensive\ndatabases of materials properties, are becoming a fruitful direction for\nmaterials analysis. As a result, a plethora of models have been constructed and\ntrained on existing data to predict properties of new systems. These powerful\nmethods allow researchers to target studies only at interesting materials\n unicodex2014 neglecting the non-synthesizable systems and those without\nthe desired properties unicodex2014 thus reducing the amount of resources\nspent on expensive computations and/or time-consuming experimental synthesis.\nHowever, using these predictive models is not always straightforward. Often,\nthey require a panoply of technical expertise, creating barriers for general\nusers. AFLOW-ML (AFLOW underline\Machine\n underline\Learning) overcomes the problem by streamlining the use\nof the machine learning methods developed within the AFLOW consortium. The\nframework provides an open RESTful API to directly access the continuously\nupdated algorithms, which can be transparently integrated into any workflow to\nretrieve predictions of electronic, thermal and mechanical properties. These\ntypes of interconnected cloud-based applications are envisioned to be capable\nof further accelerating the adoption of machine learning methods into materials\ndevelopment.\n