Condensed Matter > Disordered Systems and Neural Networks
[Submitted on 1 Aug 2018 (this version), latest version 28 Mar 2019 (v2)]
Title:Using Machine Learning for Scientific Discovery in Electronic Quantum Matter Visualization Experiments
View PDFAbstract:Essentials of the scientific discovery process have remained largely unchanged for millennia: systematic human observation of natural phenomena is used to form hypotheses that, when validated through experimentation, are generalized into established scientific theory. Today, however, we face major scientific challenges because automated scientific instrumentation and large-scale data acquisition are generating data sets of such volume and complexity as to defy human analysis. To achieve scientific discovery using these enormous empirical data sets will require new approaches. A good example is modern visualization studies of electronic quantum matter (EQM) which yield dense arrays of electronic structure images that are often astonishingly complex at the atomic-scale. Given the recent advances in machine learning (ML) studies of synthetic EQM data, the outstanding challenge now is to engage ML with real EQM data. Here we report development and training of an array of artificial neural networks (ANN) designed to recognize different types of hypothesized order hidden in EQM image-arrays. The ANNs are used to analyze a large, experimentally-derived EQM image archive, spanning a wide range of electron densities and energies, in carrier-doped cuprate Mott insulators. Remarkably, this ML system repeatedly and reliably discovers, throughout all these noisy and complex data, the features of a very specific ordered state of EQM. It discerns a lattice-commensurate, precisely four-unit-cell periodic, unidirectional, translational-symmetry-breaking state, consistent with those long predicted within strong-coupling theories of electronic liquid crystals. This marks a breakthrough in the capability of ML to achieve physics discovery and suggests broad new avenues for the scientific method.
Submission history
From: Andrej Mesaros [view email][v1] Wed, 1 Aug 2018 18:00:37 UTC (1,881 KB)
[v2] Thu, 28 Mar 2019 22:01:02 UTC (8,827 KB)
Current browse context:
cond-mat.dis-nn
Change to browse by:
References & Citations
Bibliographic and Citation Tools
Bibliographic Explorer (What is the Explorer?)
Litmaps (What is Litmaps?)
scite Smart Citations (What are Smart Citations?)
Code, Data and Media Associated with this Article
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
DagsHub (What is DagsHub?)
Gotit.pub (What is GotitPub?)
Papers with Code (What is Papers with Code?)
ScienceCast (What is ScienceCast?)
Demos
Recommenders and Search Tools
Influence Flower (What are Influence Flowers?)
Connected Papers (What is Connected Papers?)
CORE Recommender (What is CORE?)
IArxiv Recommender
(What is IArxiv?)
arXivLabs: experimental projects with community collaborators
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.