Characterizing transition-metal dichalcogenide thin-films using hyperspectral imaging and machine learning
Not Available
36 MATERIALS SCIENCE↗
SEARCH · Engineering Papers
Results for “Information Theory”
Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.
Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.
QUBO formulations for training machine learning models
Abstract Training machine learning models on classical computers is usually a time and compute intensive process. With Moore’s law nearing its inevitable end and an ever-increasing demand for large-scale data analysis using machine learning, we must leverage non-conventional computing paradigms like quantum computing to train machine learning models efficiently. Adiabatic quantum computers can approximately solve NP-hard problems, such as the quadratic unconstrained binary optimization (QUBO), faster than classical computers. Since many machine learning problems are also NP-hard, we believe adiabatic quantum computers might be instrumental in training machine learning models efficiently in the post Moore’s law era. In order to solve problems on adiabatic quantum computers, they must be formulated as QUBO problems, which is very challenging. In this paper, we formulate the training problems of three machine learning models—linear regression, support vector machine (SVM) and balanced k-means clustering—as QUBO problems, making them conducive to be trained on adiabatic quantum computers. We also analyze the computational complexities of our formulations and compare them to corresponding state-of-the-art classical approaches. We show that the time and space complexities of our formulations are better (in case of SVM and balanced k-means clustering) or equivalent (in case of linear regression) to their classical counterparts.
97 MATHEMATICS AND COMPUTING↗
An adaptive approach to machine learning for compact particle accelerators
Abstract Machine learning (ML) tools are able to learn relationships between the inputs and outputs of large complex systems directly from data. However, for time-varying systems, the predictive capabilities of ML tools degrade if the systems are no longer accurately represented by the data with which the ML models were trained. For complex systems, re-training is only possible if the changes are slow relative to the rate at which large numbers of new input-output training data can be non-invasively recorded. In this work, we present an approach to deep learning for time-varying systems that does not require re-training, but uses instead an adaptive feedback in the architecture of deep convolutional neural networks (CNN). The feedback is based only on available system output measurements and is applied in the encoded low-dimensional dense layers of the encoder-decoder CNNs. First, we develop an inverse model of a complex accelerator system to map output beam measurements to input beam distributions, while both the accelerator components and the unknown input beam distribution vary rapidly with time. We then demonstrate our method on experimental measurements of the input and output beam distributions of the HiRES ultra-fast electron diffraction (UED) beam line at Lawrence Berkeley National Laboratory, and showcase its ability for automatic tracking of the time varying photocathode quantum efficiency map. Our method can be successfully used to aid both physics and ML-based surrogate online models to provide non-invasive beam diagnostics.
97 MATHEMATICS AND COMPUTING↗
Quantum algorithm for electronic band structures with local tight-binding orbitals
While the main thrust of quantum computing research in materials science is to accurately measure the classically intractable electron correlation effects due to Coulomb repulsion, designing optimal quantum algorithms for simpler problems with well-understood solutions is a useful tactic to advance our quantum “toolbox”. With this in mind, we consider the quantum calculation of a periodic system’s single-electron band structure over a path through reciprocal space. Previous efforts have used the Variational Quantum Eigensolver algorithm to solve the energy of each band, which involves numerically optimizing the parameters of a variational quantum circuit to minimize a cost function, constructed as the expectation value of a Hamiltonian operator. Traditionally, a unique Hamiltonian operator is constructed for each k-point, so that many cost functions, each with their own parameter space, must be optimized to generate a single band. Similarly, calculating higher bands than the first has traditionally involved modifying the cost function with additional overlap terms to ensure higher-energy eigenstates are orthogonal to those of lower bands. In this paper, we adopt a direct space approach, using a novel hybrid first/second-quantized qubit mapping which allows us to construct a single Hamiltonian, and a single cost-function, suitable for solving the entire electronic band structure. In contrast to previous approaches, the k-point and the band index are selected by additional parameters in our quantum circuit, rather than through modifications to the cost function. The result is a technically and conceptually simpler approach to band structure calculations on a quantum computer. Moreover, we expect that the tools developed herein will motivate new strategies for tackling highly-correlated materials beyond the grasp of classical computing.
71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
Complexity-calibrated benchmarks for machine learning reveal when prediction algorithms succeed and mislead
Abstract Recurrent neural networks are used to forecast time series in finance, climate, language, and from many other domains. Reservoir computers are a particularly easily trainable form of recurrent neural network. Recently, a “next-generation” reservoir computer was introduced in which the memory trace involves only a finite number of previous symbols. We explore the inherent limitations of finite-past memory traces in this intriguing proposal. A lower bound from Fano’s inequality shows that, on highly non-Markovian processes generated by large probabilistic state machines, next-generation reservoir computers with reasonably long memory traces have an error probability that is at least $$\sim 60\%$$ ∼ 60 % higher than the minimal attainable error probability in predicting the next observation. More generally, it appears that popular recurrent neural networks fall far short of optimally predicting such complex processes. These results highlight the need for a new generation of optimized recurrent neural network architectures. Alongside this finding, we present concentration-of-measure results for randomly-generated but complex processes. One conclusion is that large probabilistic state machines—specifically, large $$\epsilon$$ ϵ -machines—are key to generating challenging and structurally-unbiased stimuli for ground-truthing recurrent neural network architectures.
97 MATHEMATICS AND COMPUTING↗
Leveraging prior mean models for faster Bayesian optimization of particle accelerators
Tuning particle accelerators is a challenging and time-consuming task that can be automated and carried out efficiently using suitable optimization algorithms, such as model-based Bayesian optimization techniques. One of the major advantages of Bayesian algorithms is the ability to incorporate prior information about beam physics and historical behavior into the model used to make control decisions. In this work, we examine incorporating prior accelerator physics information into Bayesian optimization algorithms by utilizing fast executing, neural network models trained on simulated or historical datasets as prior mean functions in Gaussian process models. We show that in ideal cases, this technique substantially increases convergence speed to optimal solutions in high-dimensional tuning parameter spaces. Additionally, we demonstrate that even in non-ideal cases, where prior models of beam dynamics do not exactly match experimental conditions, the use of this technique can still enhance convergence speed. Finally, we demonstrate how these methods can be used to improve optimization in practical applications, such as transferring information gained from beam dynamics simulations to online control of the LCLS injector, and transferring knowledge gained from experimental measurements across different operating modes, such as accelerating different ion species at the ATLAS heavy ion accelerator.
43 PARTICLE ACCELERATORS↗
Discovering nuclear models from symbolic machine learning
Numerous phenomenological nuclear models have been proposed to describe specific observables within different regions of the nuclear chart. However, developing a unified model that describes the complex behavior of all nuclei remains an open challenge. Here, we explore whether symbolic Machine Learning (ML) can rediscover traditional nuclear physics models or identify alternatives with improved simplicity, fidelity, and predictive power. To address this challenge, we developed a Multi-objective Iterated Symbolic Regression approach that handles symbolic regressions over multiple target observables, accounts for experimental uncertainties and is robust against high-dimensional problems. As a proof of principle, we applied this method to describe the nuclear binding energies and charge radii of light and medium mass nuclei. Our approach identified simple analytical relationships based on the number of protons and neutrons, providing interpretable models with precision comparable to state-of-the-art nuclear models. Additionally, we integrated this ML-discovered model with an existing complementary model to estimate the limits of nuclear stability. These results highlight the potential of symbolic ML to develop accurate nuclear models and guide our description of complex many-body problems.
Nuclear structure↗
Super-resolution and signal separation in contact Kelvin probe force microscopy of electrochemically active ferroelectric materials
In this work, imaging mechanisms in contact Kelvin probe force microscopy (cKPFM) are explored via information theory-based methods. Gaussian processes are used to achieve super-resolution in the cKPFM signal, effectively extrapolating across the spatial and parameter space. Tensor factorization is applied to reduce the multidimensional signal to the tensor convolution of the scalar functions that show a clear trending behavior with the imaging parameters. These methods establish a workflow for the analysis of the multidimensional datasets that can then be related to the relevant physical mechanisms. We also provide an interactive Google Colab notebook that goes through all the analyses discussed in the paper.
36 MATERIALS SCIENCE↗
ET-AL: Entropy-targeted active learning for bias mitigation in materials data
Growing materials data and data-driven informatics drastically promote the discovery and design of materials. While there are significant advancements in data-driven models, the quality of data resources is less studied despite its huge impact on model performance. In this work, we focus on data bias arising from uneven coverage of materials families in existing knowledge. Observing different diversities among crystal systems in common materials databases, we propose an information entropy-based metric for measuring this bias. To mitigate the bias, we develop an entropy-targeted active learning (ET-AL) framework, which guides the acquisition of new data to improve the diversity of underrepresented crystal systems. We demonstrate the capability of ET-AL for bias mitigation and the resulting improvement in downstream machine learning models. This approach is broadly applicable to data-driven materials discovery, including autonomous data acquisition and dataset trimming to reduce bias, as well as data-driven informatics in other scientific domains.
36 MATERIALS SCIENCE↗
Complexity analysis of a CT injection experiment on BRB
In this work, we use Jensen–Shannon complexity and permutation entropy to analyze the magnetic field fluctuations of an astrophysically scaled plasma experiment. The experiment was intended to emulate an interplanetary coronal mass ejection event in the lab, recreating the major sections seen in satellite data. We also use a technique called “delay,” in which we use select elements, skipping one or more data points at a time, in our time series data to obtain Jensen–Shannon complexity as a function of frequency and investigate the frequency of maximized complexity. We then compare the delay frequencies to other frequencies in the plasma. We found that the frequencies for maximum complexity do not correspond to the frequencies investigated, implying that other physical mechanisms lead to an increase in complexity at these frequencies.
70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
Whole-proteome tree of life suggests a deep burst of organism diversity
An organism tree of life (organism ToL) is a conceptual and metaphorical tree to capture a simplified narrative of the evolutionary course and kinship among the extant organisms. Such a tree cannot be experimentally validated but may be reconstructed based on characteristics associated with the organisms. Since the whole-genome sequence of an organism is, at present, the most comprehensive descriptor of the organism, a whole-genome sequence-based ToL can be an empirically derivable surrogate for the organism ToL. However, experimentally determining the whole-genome sequences of many diverse organisms was practically impossible until recently. We have constructed three types of ToLs for diversely sampled organisms using the sequences of whole genome, of whole transcriptome, and of whole proteome. Of the three, whole-proteome sequence-based ToL (whole-proteome ToL), constructed by applying information theory-based feature frequency profile method, an “alignment-free” method, gave the most topologically stable ToL. Here, we describe the main features of a whole-proteome ToL for 4,023 species with known complete or almost complete genome sequences on grouping and kinship among the groups at deep evolutionary levels. The ToL reveals 1) all extant organisms of this study can be grouped into 2 “Supergroups,” 6 “Major Groups,” or 35+ “Groups”; 2) the order of emergence of the “founders” of all of the groups may be assigned on an evolutionary progression scale; 3) all of the founders of the groups have emerged in a “deep burst” at the very beginning period near the root of the ToL—an explosive birth of life’s diversity.
59 BASIC BIOLOGICAL SCIENCES↗
From minimum-viable-products to full models: a step-wise development of diagnostic forward models in support of design, analysis and modelling on the ST40 tokamak
Like most magnetic confined fusion experiments, the ST40 tokamak started off with a small subset of diagnostics and gradually increased the diagnostic set to include more complex and comprehensive systems. To make the most of each operational phase, forward models of various diagnostics are used and developed to aid design, provide consistency-checks during commissioning, test analysis methods, and build workflows to constrain high-level parameters to inform interpretation, theory and modelling. For new models and new analysis workflows, minimum-viable-products are released early, and their complexity is increased in a step-wise manner, facilitating the support of all programme phases on multiple parallel applications, while enabling learning opportunities and feedback loops. In this contribution we review the philosophy, scope and architecture of the framework under development. We discuss the details of some forward models, with examples on how they are used to aid diagnostic design, to investigate analysis methodologies through synthetic data, and how they are embedded in experimental analysis workflows. We compare previously published experimental results with new, more advanced analysis workflows employing more recent, detailed models and new diagnostic data, providing confirmation of the published material from the 2021–22 experimental campaign.
integrated data analysis↗
Decoupling plasma, catalyst, and gaseous mechanisms for non-oxidative methane conversion
Direct non-oxidative methane (CH 4 ) conversion to value-added hydrogen (H 2 ) and C 2 products remains hindered by fundamental catalytic scaling constraints and rapid surface deactivation at elevated temperatures. Plasma-enabled catalysis offers a promising route to overcome the thermodynamic and kinetic barriers of direct non-oxidative methane upgrading at mild conditions, yet control over C–C product selectivity and catalyst stability remains elusive. Here, we establish a unified mechanistic framework including Langmuir–Hinshelwood (L–H) and Langmuir–Rideal (L–R) mechanisms that disentangles the roles of plasma excitation (including vibrationally activated species and radicals), surface temperature (T sur ), and catalyst binding energy in steering CH 4 conversion to H 2 and C 2 hydrocarbons. Through a combination of density functional theory (DFT) informed microkinetic modeling, in situ and ex situ surface characterization, and product quantification under dielectric barrier discharge conditions, we show that vibrationally excited CH 4 lowers activation barriers selectively for dissociative chemisorption, enabling surface activation across a wide range of transition metal catalysts at low thermal energy input. We find that once CH 4 is dissociatively chemisorbed, the branching between C 2 H 2 , C 2 H 4 , and C 2 H 6 is governed by surface properties (carbon binding energy, T sur , etc), regardless of plasma excitation. The DFT informed microkinetic model decouples the effects of molecular activation from surface properties and indentifies operating windows that maximize target yields while suppressing carbon accumulation and subsequent catalytic inactivation. Experiments on polycrystalline Cu/Al 2 O 3 , Ni/Al 2 O 3 , and Pt/Al 2 O 3 validate these predictions, revealing catalyst-dependent branching toward ethane or ethylene and distinct deactivation profiles. We unify these trends into a generalized three-dimensional plasma-thermal-catalytic design space, from which reduced descriptors such as T vib /T sur in the limit of vibrationally excited L–H pathways emerge as predictive metrics. These results enable rational tuning of methane conversion pathways and unlock selective C 2 formation using earth-abundant metals under mild plasma conditions.
catalyst inactivation↗
Tabulated equations of state from models informed by chiral effective field theory
Abstract We construct four equation of state (EoS) tables, tabulated over a range of temperatures, densities, and charge fractions, relevant for neutron star applications such as simulations of neutron star mergers. The EoS are computed from a relativistic mean-field theory constrained by the pure neutron matter EoS from chiral effective field theory, inferred properties of isospin-symmetric nuclear matter, and astrophysical observations of neutron star structure. To model nuclear matter at low densities, we attach an EoS that models inhomogeneous nuclear matter at arbitrary temperatures and charge fractions. The four EoS tables we develop are available from the CompOSE EoS repository https://compose.obspm.fr/eos/297 and gitlab.com/ahaber/qmc-rmf-tables .
79 ASTRONOMY AND ASTROPHYSICS↗
Stable isotopes contain substantial additive information about terrestrial carbon and water cycling
Abstract Stable isotope ratios of H ( δ 2 H ), O ( δ 18 O), and C ( δ 13 C) are linked to key biogeochemical processes of the water and carbon cycles; however, the degree to which isotope-associated processes are reflected in macroscale ecosystem flux observations remains unquantified. Here through formal information assessment, new measurements of δ 13 C of net ecosystem exchange ( NEE ) as well as δ 2 H and δ 18 O of latent heat ( LH ) fluxes across the United States National Ecological Observation Network (NEON) are used to determine conditions under which isotope measurements are informative of environmental exchanges. We find all three isotopic datasets individually contain comparable amounts of information about NEE and LH fluxes as wind speed observations. Such information from isotope measurements, however, is largely unique. Generally, δ 13 C provides more information about LH as aridity increases or mean annual precipitation decreases. δ 2 H provides more information about LH as temperatures or mean annual precipitation decreases, and also provides more information about NEE as temperatures decrease. Overall, we show that the stable isotope datasets collected by NEON contribute non-trivial amounts of new information about bulk environmental fluxes useful for interpreting biogeochemical and ecohydrological processes at landscape scales. However, the utility of this new information varies with environmental conditions at continental scales. This study provides an approach for quantifying the value adding non-traditional sensing approaches to environmental monitoring sites and the patterns identified here are expected to aid in modeling and data interpretation efforts focused on constraining carbon and water cycles’ mechanisms.
54 ENVIRONMENTAL SCIENCES↗
Coarse-grain cluster analysis of tensors with application to climate biome identification
A tensor provides a concise way to codify the interdependence of complex data. Treating a tensor as a d-way array, each entry records the interaction between the different indices. Clustering provides a way to parse the complexity of the data into more readily understandable information. Clustering methods are heavily dependent on the algorithm of choice, as well as the chosen hyperparameters of the algorithm. However, their sensitivity to data scales is largely unknown. In this work, we apply the discrete wavelet transform to analyze the effects of coarse-graining on clustering tensor data. We are particularly interested in understanding how scale affects clustering of the Earth's climate system. The discrete wavelet transform allows classification of the Earth's climate across a multitude of spatial-temporal scales. The discrete wavelet transform is used to produce an ensemble of classification estimates, as opposed to a single classification. Each element of the ensemble is a clustering at a different spatial-temporal scale. Information theoretic approaches are used to identify important scale lengths in clustering the L15 Climate Datset. We also discover a sub-collection of the ensemble that spans the majority of the variance observed, allowing for efficient consensus clustering techniques that can be used to identify climate biomes.
54 ENVIRONMENTAL SCIENCES↗
Exploring nonmultiplicativity in the geometric measure of entanglement
The geometric measure of entanglement (GME) quantifies how close a multipartite quantum state is to the set of separable states under the Hilbert-Schmidt inner product. The GME can be nonmultiplicative, meaning that the closest product state to two states is entangled across subsystems. In this work, we explore the GME in two families of states: those that are invariant under bilateral orthogonal (𝑂⊗𝑂) transformations, and mixtures of singlet states. In both cases, a region of GME nonmultiplicativity is identified around the antisymmetric projector state. We employ state-of-the-art numerical optimization methods and models to quantitatively analyze nonmultiplicativity in these states for 𝑑=3. Here, we also investigate a constrained form of GME that measures closeness to the set of real product states and show that this measure can be nonmultiplicative even for real separable states.
Quantum correlations in quantum information↗
Achievable Rates for Concatenated Square Gottesman-Kitaev-Preskill Codes
The Gottesman-Kitaev-Preskill (GKP) codes are known to achieve optimal rates under displacement noise and pure-loss channels, which establishes theoretical foundations for its optimality. However, such optimal rates are only known to be achieved at a discrete set of noise strengths with the current self-dual symplectic lattice construction. In this work, we develop a new coding strategy using concatenated continuous variable-discrete variable encodings to go beyond past results and establish GKP’s optimal rate over all noise strengths. In particular, for displacement noise, the rate is obtained through a constructive approach by concatenating GKP codes with a quantum polar code and analog decoding. For a pure-loss channel, we prove the existence of capacity-achieving GKP codes through a random coding approach. These results highlight the capability of concatenation-based GKP codes and provides new methods for constructing good GKP lattices.
71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗