We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cond-mat.mtrl-sci

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Condensed Matter > Materials Science

Title: Physics guided deep learning generative models for crystal materials discovery

Abstract: Deep learning based generative models such as deepfake have been able to generate amazing images and videos. However, these models may need significant transformation when applied to generate crystal materials structures in which the building blocks, the physical atoms are very different from the pixels. Naively transferred generative models tend to generate a large portion of physically infeasible crystal structures that are not stable or synthesizable. Herein we show that by exploiting and adding physically oriented data augmentation, loss function terms, and post processing, our deep adversarial network (GAN) based generative models can now generate crystal structures with higher physical feasibility and expand our previous models which can only create cubic structures.
Subjects: Materials Science (cond-mat.mtrl-sci); Machine Learning (cs.LG)
Journal reference: AAAI Fall Symposium Series (FSS) 2021
Cite as: arXiv:2112.03528 [cond-mat.mtrl-sci]
  (or arXiv:2112.03528v1 [cond-mat.mtrl-sci] for this version)

Submission history

From: Jianjun Hu [view email]
[v1] Tue, 7 Dec 2021 06:54:48 GMT (8188kb,D)

Link back to: arXiv, form interface, contact.