Correlator convolutional neural networks as an interpretable architecture for image-like quantum matter data

Abstract

Image-like data from quantum systems promises to offer greater insight into the physics of correlated quantum matter. However, the traditional framework of condensed matter physics lacks principled approaches for analyzing such data. Machine learning models are a powerful theoretical tool for analyzing image-like data including many-body snapshots from quantum simulators. Recently, they have successfully distinguished between simulated snapshots that are indistinguishable from one and two point correlation functions. Thus far, the complexity of these models has inhibited new physical insights from such approaches. Here, we develop a set of nonlinearities for use in a neural network architecture that discovers features in the data which are directly interpretable in terms of physical observables. Applied to simulated snapshots produced by two candidate theories approximating the doped Fermi-Hubbard model, we uncover that the key distinguishing features are fourth-order spin-charge correlators. Our approach lends itself well to the construction of simple, versatile, end-to-end interpretable architectures, thus paving the way for new physical insights from machine learning studies of experimental and numerical data.

Analysis of Ice and Snow Path Planning System Based on MNN Algorithm

“…Grattarola et al put forward a path optimization planning model based on firefly tracking algorithm according to the specific ice and snow movement path law, which has the advantages of strong stability and good reliability [ 14 ]. According to the characteristics of ice and snow movement path, Miles et al improved the path planning strategy and proposed an intelligent path planning method based on super-high selection [ 15 ]. The research of Janson et al shows that different types of path optimization strategies can be adopted according to the differences of ice and snow movement; that is, the differences in different path planning schemes can be quickly identified first, and then the common path planning standard schemes can be analyzed point-to-point, and the analysis result data can be converted to the path signal, and the final path planning scheme can be output [ 16 ].…”

Modeling of a Generic Edge Computing Application Design

“…In this sense, some enhancements of CNN have been proposed for specific duties in recent times, to obtain greater accuracy in predicting visual recognition in data science, such as subpixel displacement measures [36], defect identification in high-speed trains [37], correlating image-like data out of quantum systems [38], modeling wind field downscaling [39], designing a zero knowledge proof scheme [40], classifying satellite image time series [41], working with ensembles [42], dealing with osteoporosis diagnoses [43], screening and staging diabetic retinopathy [44], analyzing cloud particles [45], inspecting diffraction data [46], or examining x-ray images [47].…”

Unsupervised learning of two-component nematicity from STM data on magic angle bilayer graphene

“…Here, we propose an unsupervised machine learning approach to detect the two types of nematicity and their spatial fluctuations from STM data on moiré materials. Lately, the community has made rapid progress in applying machine learning [26? ] to experimental data on quantum matter from bulk probes such as resonant ultrasound [27], neutron scattering [28], and X-ray scattering [29] and microscopic probes such as STM [30][31][32], electron microscopy [33], and quantum gas microscopy [34][35][36]. However, much of the literature has focused on supervised machine learning.…”