A perspective on machine learning and data science for strongly correlated electron problems

Johnston, S.; Khatami, Ehsan; Scalettar, Richard

doi:10.1016/j.cartre.2022.100231

Cited by 8 publications

(4 citation statements)

References 185 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This design allows users to implement straightforward parallelization at the script level, adapt the package to their existing workflows, and more readily interface SmoQyDQMC.jl with the rich ecosystem of scientific computing packages being actively developed in the Julia programming language. For example, it can be readily coupled to existing machine learning and artificial intelligence packages to enable new research in this direction [94].…”

Section: Background and Motivationmentioning

confidence: 99%

SmoQyDQMC.jl: A flexible implementation of determinant quantum Monte Carlo for Hubbard and electron-phonon interactions

Cohen-Stead,

Costa,

Neuhaus

et al. 2024

SciPost Phys. Codebases

Self Cite

View full text Add to dashboard Cite

We introduce the SmoQyDQMC.jl package, a Julia implementation of the determinant quantum Monte Carlo algorithm. SmoQyDQMC.jl supports generalized tight-binding Hamiltonians with on-site Hubbard and generalized electron-phonon (ee-ph) interactions, including non-linear ee-ph coupling and anharmonic lattice potentials. Our implementation uses hybrid Monte Carlo methods with exact forces for sampling the phonon fields, enabling efficient simulation of low-energy phonon branches, including acoustic phonons. The SmoQyDQMC.jl package also uses a flexible scripting interface, allowing users to adapt it to different workflows and interface with other software packages in the Julia ecosystem. The code for this package can be downloaded from our GitHub repository at https://github.com/SmoQySuite/SmoQyDQMC.jl or installed using the Julia package manager. The online documentation, including examples, can be obtained from our document page at https://smoqysuite.github.io/SmoQyDQMC.jl/stable/.

show abstract

Section: Background and Motivationmentioning

confidence: 99%

SmoQyDQMC.jl: A flexible implementation of determinant quantum Monte Carlo for Hubbard and electron-phonon interactions

Cohen-Stead,

Costa,

Neuhaus

et al. 2024

SciPost Phys. Codebases

Self Cite

View full text Add to dashboard Cite

show abstract

“…The focus is on 2D magnetic systems that often contain transition metals, which belong to strongly correlated electron systems. [ 89 ] To ensure accuracy in the calculations, it is important to take into account the role of the Hubbard parameter U , which represents the Coulomb repulsion arising from strong electron correlations. [ 90 ] Among them, the U value of the transition metal can be determined by comparing experimental values of bandgaps, structural parameters, and magnetic moments.…”

Section: Framework Of Htp Screening and Adaptive MLmentioning

confidence: 99%

Advancements in High‐Throughput Screening and Machine Learning Design for 2D Ferromagnetism: A Comprehensive Review

Xin,

Song,

Jin

et al. 2023

Advcd Theory and Sims

View full text Add to dashboard Cite

Abstract2D intrinsic magnetic materials possess unique physical properties distinct from bulk materials, providing an ideal research platform for the development of low‐dimensional spintronics. The traditional approach to developing new materials involves a “trial‐and‐error” method, which is inherently flawed due to long development cycles and high costs. In recent years, with the rapid improvement in computational power, the high throughput (HTP) first‐principles calculation based on density functional theory (DFT) and machine learning (ML) method have provided a highly effective means for the design of novel intrinsic ferromagnetic materials and the study of their magnetic properties. This article reviews the recent research progress in 2D ferromagnetic materials, with particular emphasis on the significant role played by HTP first‐principles calculations and ML in the exploration and fabrication of two‐dimension ferromagnetic (2DFM) materials. Finally, the future development and challenges of 2DFM materials are discussed.

show abstract

“…These patches are subsequently linearly transformed, i.e., d−dimensional representations of the input patches, called tokens, are computed. 6 Importantly, as transformers do not sequentially process the input, the tokens are further positionally encoded, i.e., the position of the patch within the original image is stored. Thereafter follows the self-attention encoder, where all-to-all inter-dependencies between tokens are computed.…”

Section: Transformer Encodermentioning

confidence: 99%

“…Next to revolutionizing applications in image and sequence processing, in recent years neural networks have gained tremendous interest also in the field of quantum many-body physics [4][5][6][7]. In strongly correlated systems, complex phases of matter can emerge in seemingly simple models -which, in many settings, still lack microscopic understanding [8,9].…”

Section: Introductionmentioning

confidence: 99%

Fluctuation based interpretable analysis scheme for quantum many-body snapshots

Schlömer,

Bohrdt

2023

SciPost Phys.

View full text Add to dashboard Cite

Microscopically understanding and classifying phases of matter is at the heart of strongly-correlated quantum physics. With quantum simulations, genuine projective measurements (snapshots) of the many-body state can be taken, which include the full information of correlations in the system. The rise of deep neural networks has made it possible to routinely solve abstract processing and classification tasks of large datasets, which can act as a guiding hand for quantum data analysis. However, though proven to be successful in differentiating between different phases of matter, conventional neural networks mostly lack interpretability on a physical footing. Here, we combine confusion learning [Nat. Phys. 13, 435 (2017)] with correlation convolutional neural networks [Nat. Commun. 12, 3905 (2021)], which yields fully interpretable phase detection in terms of correlation functions. In particular, we study thermodynamic properties of the 2D Heisenberg model, whereby the trained network is shown to pick up qualitative changes in the snapshots above and below a characteristic temperature where magnetic correlations become significantly long-range. We identify the full counting statistics of nearest neighbor spin correlations as the most important quantity for the decision process of the neural network, which go beyond averages of local observables. With access to the fluctuations of second-order correlations — which indirectly include contributions from higher order, long-range correlations — the network is able to detect changes of the specific heat and spin susceptibility, the latter being in analogy to magnetic properties of the pseudogap phase in high-temperature superconductors [Phys. Rev. Lett. 62, 957 (1989)]. By combining the confusion learning scheme with transformer neural networks, our work opens new directions in interpretable quantum image processing being sensible to long-range order.

show abstract

A perspective on machine learning and data science for strongly correlated electron problems

Cited by 8 publications

References 185 publications

SmoQyDQMC.jl: A flexible implementation of determinant quantum Monte Carlo for Hubbard and electron-phonon interactions

SmoQyDQMC.jl: A flexible implementation of determinant quantum Monte Carlo for Hubbard and electron-phonon interactions

Advancements in High‐Throughput Screening and Machine Learning Design for 2D Ferromagnetism: A Comprehensive Review

Fluctuation based interpretable analysis scheme for quantum many-body snapshots

Contact Info

Product

Resources

About