Search NASA⌕ Search

SEARCH · Search NASA

Results for “Random effect model”

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.

At least 1,513 records · Page 84

Probabilistic Modeling of a Three-Stage Human Landing System Architecture

Space Policy Directive-1 has led to NASA partnerships with commercial entities on procurement which includes the development of the Human Landing System (HLS) [1]. With the goal of delivering human crew to the lunar surface by 2024, system uncertainties become an important obstacle to the maturation of multiple new, driving technologies and mission concepts of the HLS program. As unmitigated uncertainties have previously led to failed development programs, these risks and their impacts must be understood and handled to ensure program success [2]. Sources of uncertainty include novel engine designs and configurations, increased reliance on cryogenic fluid management(CFM), and refueling technologies—which propagate as high-level performance metrics such as overall propellant mass and engine performance. Also, the occurrence of operational uncertainties—e.g. launch conditions or need to abort during the mission—can cause cascading effects on the rest of the mission that are difficult to definitively quantify, and are outside the scope of control. These concrete examples and other occurrences can be categorized as either epistemic or aleatory uncertainties.Epistemic uncertainty arises due to a lack of knowledge and can be alleviated with design and program maturation. Aleatory uncertainty is due to the inherent randomness of the system and cannot be directly reduced, unlike epistemic uncertainty. Robust design and probabilistic methods can compensate for aleatory effects. A taxonomy of uncertainty is referred to for this work [3]. In this paper, a probabilistic methodology to handle uncertainties has been demonstrated on a three-element HLS concept [1, 4], which allows tracking of current best estimates of the concept and assessment of concept design robustness against uncertainties. A sample case has been completed for this abstract, and an expansion on the methodology will be included in the final paper. This methodology has two key parts: first, the creation of a dynamic architecture model of a three-element HLS concept; and second, its use with surrogate modeling and range estimating techniques to capture and propagate uncertainties. This abstract will cover the basics of the approach used, and further details and justifications will be in the final paper.The mission profile associated with this three-element concept (Fig 1) was modeled as a set of mission events that facilitated mass changes, idles, or spacecraft maneuvers. The mission profile scope starts with each element’s NRHO orbit insertion and aggregation and ends at post-sortie rendezvous with Orion. More detail on the mission profile will be in the final paper. The DYnamic Rocket EQuation Tool (DYREQT), a space systems synthesis and sizing framework used by NASA, was used as the physics framework to model the HLS architecture for applying the probabilistic methodology [5, 6]. Specifically, a parametric representation of the lander, ascent, and transfer elements and the mission profile of each element was established, with vehicle and mission parameters available as inputs to allow for a dynamic model. Each vehicle stage was modeled with high-level performance metrics, using Isp and propellant mass fraction (PMF) to remain parametric. For the probabilistic analysis, uncertainties of interest within the HLS concept were enumerated and represented as parameters within the DYREQT model as inputs for vehicle stages or mission profile events. These parameters were frozen at their nominal values for the purposes of baselining architecture performance and sizing the vehicle appropriately based on reference documentation [1]. Range estimating—a probabilistic method that combines Monte Carlo sampling, focus on critical parameters, and heuristics to assess risk and opportunities—is traditionally used with Mass Equipment Lists (MELs), but has been adapted with operational parameters as well as vehicle parameters in theDYREQT model to capture mission uncertainty alongside vehicle uncertainty [7, 3]. This method was selected due to its application and insight on a system from a bottom-up perspective, independence from historical rules of thumb, and ability to generate sensitivities based on design decisions and uncertainties. As a sample case for the abstract, the boiloff rates of the vehicle elements and the loiter times during the mission (simulating launch time variations and changing window of opportunities) were used with range estimating to provide preliminary results. To perform the range estimation portion of this methodology (depicted in Fig. 3, further details in final paper), the DYREQT model was sampled using a Design of Experiments (DoE) to efficiently explore the architecture design space with respect to the sample set of uncertainty parameters; 5,000 cases via Latin Hypercube Sampling were computed on the DYREQT architecture model. Then, the results were used to create surrogate models, multivariate regressions that can visualize hypercube trends in the design space, of the architecture with respect to the uncertainty parameters. Range estimating was applied to the surrogates instead of the actual models, which saves computational expense due to the bulk of cases needed for the Monte Carlo simulation as part of range estimating. Uncertainty parameters were sampled independently from triangular distributions using the DoE ranges as ‘min’ and ‘max’, and the nominal value as ‘most likely’. Based engineering intuition, some uncertainty parameters are correlated—e.g. if the main propellant has a high boil-off rate, the oxidizer should follow suit as both are related to CFM technology.While a Monte Carlo simulation samples all inputs as independent, the results would show model correlations; thus, it is efficient to sample the inputs as correlated. Using a correlation matrix constructed for the uncertainty parameters, previously independent samples were transformed to perform a Correlated Monte Carlo. A table for the DoE ranges and probability distribution parameters is shown in Table 1, and more details on Correlated Monte Carlo Simulations will be discussed in the final paper. The model’s resulting DoE showed that multivariate polynomial equations fit via least squares method captured its behavior accurately for the sample case. For the Correlated Monte Carlo Simulation, a positive correlation between fuel and oxidizer boiloff rates was used as a demonstration. 10,000 cases were computed with the surrogates and the launched masses for each vehicle element was collated. The results can be displayed in a probability density function (PDF), showing the impact of the uncertainty parameters chosen. Integrating the PDFs will yield a cumulative distribution function (CDF) that shows the cumulative probability of a given value on the x-axis. For the sample case, the elements’ launch mass margin was calculated and represented in as CDFs, as a demonstrated representation of figures of merit for the HLS concept. For the lander and ascent elements, the NRHO mass insertion limit is 16t; the transfer element has a limit of 30t [1]. It can be seen with Figure 2 that this probabilistic methodology can provide insight into mass margin with respect to the uncertainties being modeled. Currently, the results show that the lander (descent) vehicle element has the most restrictive design space; it is the only element to show a 10% probability of negative margin. Further analysis on the Monte Carlo results will show sensitivities for driving constraints and parameters for architecture feasibility, which can lead to establishing potential mission rules.The combination of range estimating with a parametric architecture model for HLS demonstrated the capability of this probabilistic methodology in a sample case. As the HLS development progresses, this methodology has the potential for keeping current best estimates of architecture performance for awarded concepts due to the flexibility in DYREQT’s modeling framework and its parametric nature. Concept maturation and increased epistemic knowledge can be injected into the model probabilistic modeling, and thus continue to track probability of mission success.

Stephanie Y Zhu↗

Self-Consistent Implementation of a Zero-Equation Transport Model Into a Predictive Model for a Hall Effect Thruster

The performance of an axisymmetric multi-fluid Hall thruster code that incorporates a self-consistent, data-driven closure model for the anomalous electron transport is investigated. Five different operating conditions of the H9 magnetically shielded Hall thruster are simulated with the Jet Propulsion Laboratory’s Hall2De. In order to capture the inherent uncertainty associated with the closure model, O(100) simulations are run for each condition, each of using a coefficient set sampled randomly from a probability distribution. The results of these simulations provide probabilistic predictions of thruster performance quantities including thrust, and discharge current, as well as several component efficiencies and centerline plasma properties. The model is found to yield converged solutions at all conditions, with large 10 kHzrange oscillations and performance trends with voltage and flow rate similar to experiment. The model under-predicts the thrust by 15-25% and over-predicts the discharge current by 20% on average compared to experiments at the same discharge voltage and mass flow rate. This performance discrepancy is due to lower beam utilization, mass utilization, and divergence efficiency than experiment, resulting from high Hall parameters in the acceleration region, which lead to a protracted ion acceleration region. The physical processes underlying this result are discussed in the context of future data-driven modeling efforts.

Jorns, Benjamin A.↗

Commensurability effects in one-dimensional Anderson localization: Anomalies in eigenfunction statistics

Highlights: > Statistics of normalized eigenfunctions in one-dimensional Anderson localization at E = 0 is studied. > Moments of inverse participation ratio are calculated. > Equation for generating function is derived at E = 0. > An exact solution for generating function at E = 0 is obtained. > Relation of the generating function to the phase distribution function is established. - Abstract: The one-dimensional (1d) Anderson model (AM), i.e. a tight-binding chain with random uncorrelated on-site energies, has statistical anomalies at any rational point f=(2a)/({lambda}{sub E}) , where a is the lattice constant and {lambda}{sub E} is the de Broglie wavelength. We develop a regular approach to anomalous statistics of normalized eigenfunctions {psi}(r) at such commensurability points. The approach is based on an exact integral transfer-matrix equation for a generating function {Phi}{sub r}(u, {phi}) (u and {phi} have a meaning of the squared amplitude and phase of eigenfunctions, r is the position of the observation point). This generating function can be used to compute local statistics of eigenfunctions of 1d AM at any disorder and to address the problem of higher-order anomalies at f=p/q with q > 2. The descender of the generating function P{sub r}({phi}){identical_to}{Phi}{sub r}(u=0,{phi}) is shown to be the distribution function of phase which determines the Lyapunov exponent and the local density of states. In the leading order in the small disorder we derived a second-order partial differential equation for the r-independent ('zero-mode') component {Phi}(u, {phi}) at the E = 0 (f=1/2 ) anomaly. This equation is nonseparable in variables u and {phi}. Yet, we show that due to a hidden symmetry, it is integrable and we construct an exact solution for {Phi}(u, {phi}) explicitly in quadratures. Using this solution we computed moments I{sub m} = N< vertical bar {psi} vertical bar {sup 2m}> (m {>=} 1) for a chain of the length N {yields} {infinity} and found an essential difference between their m-behavior in the center-of-band anomaly and for energies outside this anomaly. Outside the anomaly the 'extrinsic' localization length defined from the Lyapunov exponent coincides with that defined from the inverse participation ratio ('intrinsic' localization length). This is not the case at the E = 0 anomaly where the extrinsic localization length is smaller than the intrinsic one. At E = 0 one also observes an anomalous enhancement of large moments compatible with existence of yet another, much smaller characteristic length scale.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Strong-field Gravitational Effects in the Timing and Spectroscopy of X-Ray Binaries

Using the ray-tracing code described, we developed additional features to the hot spot model to explain the broadening of the quasi-periodic oscillation (QPO) peaks as well as the damping of higher frequency harmonics in the power spectrum. We derived a number of analytic results that agree closely with more detailed numerical calculations. Three primary pieces are developed: the superposition of multiple hot spots with finite lifetimes and random phases, a width in the radial distribution of geodesic orbits, and the scattering of photons from the hot spot through a corona around the black hole. The complete model is used to fit the observed power spectra of both type A and type B QPOs seen in XTE 51550-564, giving confidence limits on each of the model parameters. These results were presented in a conference paper and in a refereed publication. We also explored the possibility of using higher-order statistics, particularly the bispectrum and bicoherence, to distinguish between competing QPO models. These methods were applied to the hot spot model and a method was proposed to map out the spacetime metric around Kerr black holes in Ref. 3 below.

Bertschinger, Edmund↗

Online technique for detecting state of onboard fiber optic gyroscope

Although angle random walk (ARW) of fiber optic gyroscope (FOG) has been well modeled and identified before being integrated into the high-accuracy attitude control system of satellite, aging and unexpected failures can affect the performance of FOG after launch, resulting in the variation of ARW coefficient. Therefore, the ARW coefficient can be regarded as an indicator of “state of health” for FOG diagnosis in some sense. The Allan variance method can be used to estimate ARW coefficient of FOG, however, it requires a large amount of data to be stored. Moreover, the procedure of drawing slope lines for estimation is painful. To overcome the barriers, a weighted state-space model that directly models the ARW to obtain a nonlinear state-space model was established for FOG. Then, a neural extended-Kalman filter algorithm was implemented to estimate and track the variation of ARW in real time. The results of experiment show that the proposed approach is valid to detect the state of FOG. Moreover, the proposed technique effectively avoids the storage of data.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

An effective field theory for non-maximal quantum chaos

In non-maximally quantum chaotic systems, the exponential behavior of out-of-time-ordered correlators (OTOCs) results from summing over exchanges of an infinite tower of higher “spin” operators. We construct an effective field theory (EFT) to capture these exchanges in (0 + 1) dimensions. The EFT generalizes the one for maximally chaotic systems, and reduces to it in the limit of maximal chaos. The theory predicts the general structure of OTOCs both at leading order in the 1/N expansion (N is the number of degrees of freedom), and after resuming over an infinite number of higher order 1/N corrections. These general results agree with those previously explicitly obtained in specific models. We also show that the general structure of the EFT can be extracted from the large q SYK model.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Machine learning in materials research: Developments over the last decade and challenges for the future

The number of studies that apply machine learning (ML) to materials science has been growing at a rate of approximately 1.67 times per year over the past decade. In this review, I examine this growth in various contexts. First, I present an analysis of the most commonly used tools (software, databases, materials science methods, and ML methods) used within papers that apply ML to materials science. The analysis demonstrates that despite the growth of deep learning techniques, the use of classical machine learning is still dominant as a whole. It also demonstrates how new research can effectively build upon past research, particular in the domain of ML models trained on density functional theory calculation data. Next, I present the progression of best scores as a function of time on the matbench materials science benchmark for formation enthalpy prediction. In particular, a dramatic improvement of 7 times reduction in error is obtained when progressing from feature-based methods that use conventional ML (random forest, support vector regression, etc.) to the use of graph neural network techniques. Finally, I provide views on future challenges and opportunities, focusing on data size and complexity, extrapolation, interpretation, access, and relevance.

36 MATERIALS SCIENCE↗

Design, Synthesis, and Self-Assembly of Polymers with Tailored Graft Distributions

Grafting density and graft distribution impact the chain dimensions and physical properties of polymers. However, achieving precise control over these structural parameters presents long-standing synthetic challenges. In this report, we introduce a versatile strategy to synthesize polymers with tailored architectures via grafting-through ring-opening metathesis polymerization (ROMP). One-pot copolymerization of an ω-norbornenyl macromonomer and a discrete norbornenyl co-monomer (diluent) provides opportunities to control the backbone sequence and therefore the side chain distribution. Toward sequence control, the homopolymerization kinetics of 23 diluents were studied, representing diverse variations in the stereochemistry, anchor groups, and substituents. These modifications tuned the homopolymerization rate constants over two orders of magnitude (0.36 M -1 s -1 < k homo < 82 M -1 s -1 ). Rate trends were identified and elucidated by complementary mechanistic and density functional theory (DFT) studies. Building on this foundation, complex architectures were achieved through copolymerizations of selected diluents with a poly (D,L-lactide) (PLA), polydimethylsiloxane (PDMS), or polystyrene (PS) macromonomer. The cross-propagation rate constants were obtained by non-linear least squares fitting of the instantaneous co-monomer concentrations according to the Mayo-Lewis terminal model. Indepth kinetic analyses indicate a wide range of accessible macromonomer/diluent reactivity ratios (0.08 < r 1 /r 2 < 20), corresponding to blocky, gradient, or random backbone sequences. We further demonstrated the versatility of this copolymerization approach by synthesizing AB graft diblock polymers with tapered, uniform, and inverse-tapered molecular “shapes.” Small-angle X-ray scattering analysis of the self-assembled structures illustrates effects of the graft distribution on the domain spacing and backbone conformation. Collectively, the insights provided herein into the ROMP mechanism, monomer design, and homo- and copolymerization rate trends offer a general strategy for the design and synthesis of graft polymers with arbitrary architectures. Lastly, controlled copolymerization therefore expands the parameter space for molecular and materials design.

36 MATERIALS SCIENCE↗

Coupling flux balance analysis with reactive transport modeling through machine learning for rapid and stable simulation of microbial metabolic switching

Integrating genome-scale metabolic networks with reactive transport models (RTMs) provides a detailed description of the dynamic changes in microbial growth and metabolism. Despite promising demonstrations in the past, computational inefficiency has been pointed out as a critical issue to overcome because it requires repeated application of linear programming (LP) to obtain flux balance analysis (FBA) solutions in every time step and spatial grid. To address this challenge, we propose a new simulation method where we train and validate artificial neural networks (ANNs) using randomly sampled FBA solutions and incorporate the resulting surrogate FBA model (represented as algebraic equations) into RTMs as source/sink terms. We demonstrate the efficiency of our method via a case study of Shewanella oneidensis MR-1. During aerobic growth on lactate, S. oneidensis produces metabolic byproducts (such as pyruvate and acetate), which are subsequently consumed as alternative carbon sources when the preferred nutrients are depleted. To effectively simulate these complex dynamics, we used a cybernetic approach that models metabolic switches as the outcome of dynamic competition among multiple growth options. In both zero-dimensional batch and one-dimensional column configurations, the ANN-based surrogate models achieved substantial reduction of computational time by several orders of magnitude compared to the original LP-based FBA models. Moreover, the ANN models produced robust solutions without any special measures to prevent numerical instability. These developments significantly promote our ability to utilize genome-scale networks in complex, multi-physics, and multi-dimensional ecosystem modeling.

59 BASIC BIOLOGICAL SCIENCES↗

Latest developments in seismic texture analysis for subsurface structure, facies, and reservoir characterization: A review

In exploration geology and geophysics, seismic texture is still a developing concept that has not been sufficiently known, although quite a number of different algorithms have been published in the literature. This paper provides a review of the seismic texture concepts and methodologies, focusing on latest developments in seismic amplitude texture analysis, with particular reference to the gray level co-occurrence matrix (GLCM) and the texture model regression (TMR) methods. The GLCM method evaluates spatial arrangements of amplitude samples within an analysis window using a matrix (a two-dimensional histogram) of amplitude co-occurrence. The matrix is then transformed into a suite of texture attributes, such as homogeneity, contrast, and randomness, which provide the basis for seismic facies classification. The TMR method uses a texture model as reference to discriminate among seismic features based on a linear, least-squares regression analysis between the model and the data within an analysis window. By implementing customized texture model schemes, the TMR algorithm has the flexibility to characterize subsurface geology for different purposes. A texture model with a constant phase is effective at enhancing the visibility of seismic structural fabrics, a texture model with a variable phase is helpful for visualizing seismic facies, and a texture model with variable amplitude, frequency, and size is instrumental in calibrating seismic to reservoir properties. Preliminary test case studies in the very recent past have indicated that the latest developments in seismic texture analysis have added to the existing amplitude interpretation theories and methodologies. These and future developments in seismic texture theory and methodologies will hopefully lead to a better understanding of the geologic implications of the seismic texture concept and to an improved geologic interpretation of reflection seismic amplitude

58 GEOSCIENCES↗

Impact of Different Oseltamivir Regimens on Treating Influenza A Virus Infection and Resistance Emergence: Insights from a Modelling Study

Several studies have proven oseltamivir to be efficient in reducing influenza viral titer and symptom intensity. However, the usefulness of oseltamivir can be compromised by the emergence and spread of drug-resistant virus. The selective pressure exerted by different oseltamivir therapy regimens have received little attention. Combining models of drug pharmacokinetics, pharmacodynamics, viral kinetics and symptom dynamics, we explored the efficacy of oseltamivir in reducing both symptoms (symptom efficacy) and viral load (virological efficacy). We simulated samples of 1000 subjects using previously estimated between-subject variability in viral and symptom dynamic parameters to describe the observed heterogeneity in a patient population. We simulated random mutations conferring resistance to oseltamivir. We explored the effect of therapy initiation time, dose, intake frequency and therapy duration on influenza infection, illness dynamics, and emergence of viral resistance. Symptom and virological efficacies were strongly associated with therapy initiation time. The proportion of subjects shedding resistant virus was 27-fold higher when prophylaxis was initiated during the incubation period compared with no treatment. It fell to below 1% when treatment was initiated after symptom onset for twice-a-day intakes. Lower doses and prophylaxis regimens led to lower efficacies and increased risk of resistance emergence. We conclude that prophylaxis initiated during the incubation period is the main factor leading to resistance emergence.

59 BASIC BIOLOGICAL SCIENCES↗

A Code to Code Benchmark for HTGR Fuel Element Depletion

The current lack of well-qualified experimental measurements for spent HGTR fuel elements limits the validation of advanced doubly heterogeneous depletion methods. Because of this shortage of data, a benchmark was developed as the first, simplest phase in a planned series of increasingly complex set of code-to-code benchmarks. The intent of this benchmark was to encourage submission of a wide range of computational results for depletion calculations in a set of basic fuel cell models. Comparison of results using independent methods and data should provide insight into potential limitations in various modeling approximations. The benchmark seeks to provide the simplest possible models, in order to minimize the effect of competing and potentially offsetting phenomena that might mask weaknesses in given methods. Twenty-one sets of results were submitted by twelve participants internationally. Various approaches were used for the treatment of microscopic grains dispersed in fuel elements in the second and third parts of the benchmark, including explicit treatment of grains distributed within a regular lattice, a randomized distribution of explicit grains, reactivity-equivalent physical transformation and a two-pass homogenization with disadvantage factors. The broad agreement appears to validate each of the double heterogeneity treatments.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗

Radiative transfer modelling of discontinuous tree canopies at microwave frequencies

Development of a first-order radiative transfer model for predicting backscatter from tree canopies has been underway at the University of Michigan Radiation Laboratory for some time. This model is known as the Michigan Microwave Canopy Scattering (MIMICS) model. This article presents the second-generation MIMICS model (MIMICS II) which accounts for canopies with discontinuous (open) crown layer geometries. MIMICS II models open crown layers by treating the location, size, and shape of the individual tree crowns as random variables. The backscattering coefficients for the canopy are then determined by introducing statistics derived from these parameters into the radiative transfer solution. Application of the radiative transfer equations to the discontinuous canopy geometry is presented. The resulting model is a robust fully polarimetric solution that is applicable over a wide variety of canopy architectures. Model simulations are compared to results generated with the continuous canopy model. The effect of the open crown geometry is found to be most significant at shallow incidence angles and at high frequencies for trees with well-developed crowns. Under these conditions, the gaps in the crown layer give rise to a notable increase in crown layer transmissivity which allows the radar to see through to the lower layers of the canopy more easily, thereby directly affecting the backscatter contribution of the trunks and ground.

Mcdonald, K. C.↗

Amino Acid Encoding for Deep Learning Applications

Background: The number of applications of deep learning algorithms in bioinformatics is increasing as they usually achieve superior performance over classical approaches, especially, when bigger training datasets are available. In deep learning applications, discrete data, e.g. words or n-grams in language, or amino acids or nucleotides in bioinformatics, are generally represented as a continuous vector through an embedding matrix. Recently, learning this embedding matrix directly from the data as part of the continuous iteration of the model to optimize the target prediction – a process called ‘end-to-end learning’ – has led to state-of-the-art results in many fields. Although usage of embeddings is well described in the bioinformatics literature, the potential of end-to-end learning for single amino acids, as compared to more classical manually-curated encoding strategies, has not been systematically addressed. To this end, we compared classical encoding matrices, namely one-hot, VHSE8 and BLOSUM62, to end-to-end learning of amino acid embeddings for two different prediction tasks using three widely used architectures, namely recurrent neural networks (RNN), convolutional neural networks (CNN), and the hybrid CNN-RNN. Results: By using different deep learning architectures, we show that end-to-end learning is on par with classical encodings for embeddings of the same dimension even when limited training data is available, and might allow for a reduction in the embedding dimension without performance loss, which is critical when deploying the models to devices with limited computational capacities. We found that the embedding dimension is a major factor in controlling the model performance. Surprisingly, we observed that deep learning models are capable of learning from random vectors of appropriate dimension. Conclusion: Our study shows that end-to-end learning is a flexible and powerful method for amino acid encoding. Further, due to the flexibility of deep learning systems, amino acid encoding schemes should be benchmarked against random vectors of the same dimension to disentangle the information content provided by the encoding scheme from the distinguishability effect provided by the scheme.

Deep-learning↗

Amino Acid Encoding for Deep Learning Applications

Background: The number of applications of deep learning algorithms in bioinformatics is increasing as they usually achieve superior performance over classical approaches, especially, when bigger training datasets are available. In deep learning applications, discrete data, e.g. words or n-grams in language, or amino acids or nucleotides in bioinformatics, are generally represented as a continuous vector through an embedding matrix. Recently, learning this embedding matrix directly from the data as part of the continuous iteration of the model to optimize the target prediction – a process called ‘end-to-end learning’ – has led to state-ofthe-art results in many fields. Although usage of embeddings is well described in the bioinformatics literature, the potential of end-to-end learning for single amino acids, as compared to more classical manually-curated encoding strategies, has not been systematically addressed. To this end, we compared classical encoding matrices, namely one-hot, VHSE8 and BLOSUM62, to end-to-end learning of amino acid embeddings for two different prediction tasks using three widely used architectures, namely recurrent neural networks (RNN), convolutional neural networks (CNN), and the hybrid CNN-RNN. Results: By using different deep learning architectures, we show that end-to-end learning is on par with classical encodings for embeddings of the same dimension even when limited training data is available, and might allow for a reduction in the embedding dimension without performance loss, which is critical when deploying the models to devices with limited computational capacities. We found that the embedding dimension is a major factor in controlling the model performance. Surprisingly, we observed that deep learning models are capable of learning from random vectors of appropriate dimension. Conclusion: Our study shows that end-to-end learning is a flexible and powerful method for amino acid encoding. Further, due to the flexibility of deep learning systems, amino acid encoding schemes should be benchmarked against random vectors of the same dimension to disentangle the information content provided by the encoding scheme from the distinguishability effect provided by the scheme.

Hesham ElAbd↗

Quasiparticle and quasihole states of nuclei around {sup 56}Ni

The single-particle spectral function of {sup 56}Ni has been computed within the framework of self-consistent Green's functions theory. The Faddeev random phase approximation method and the G matrix technique are used to account for the effects of long- and short-range physics on the spectral distribution. Large-scale calculations have been performed in spaces including up to ten oscillator shells. The chiral N{sup 3}LO interaction is used together with a monopole correction that accounts for eventual missing three-nucleon forces. The single-particle energies associated with nucleon transfer to valence 1p0f orbits are found to be almost converged with respect to both the size of the model space and the oscillator frequency. The results support that {sup 56}Ni is a good doubly magic nucleus. The absolute spectroscopic factors to the valence states on A=55,57 are also obtained. For the transition between the ground states of {sup 57}Ni and {sup 56}Ni, the calculations nicely agree with heavy-ion knockout experiments.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Finite-density effects in the Fredrickson-Andersen and Kob-Andersen kinetically-constrained models

We calculate the corrections to the thermodynamic limit of the critical density for jamming in the Kob-Andersen and Fredrickson-Andersen kinetically-constrained models, and find them to be finite-density corrections, and not finite-size corrections. We do this by introducing a new numerical algorithm, which requires negligible computer memory since contrary to alternative approaches, it generates at each point only the necessary data. The algorithm starts from a single unfrozen site and at each step randomly generates the neighbors of the unfrozen region and checks whether they are frozen or not. Our results correspond to systems of size greater than 10{sup 7} × 10{sup 7}, much larger than any simulated before, and are consistent with the rigorous bounds on the asymptotic corrections. We also find that the average number of sites that seed a critical droplet is greater than 1.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Effects of Eddy Viscosity on Time Correlations in Large Eddy Simulation

Subgrid-scale (SGS) models for large. eddy simulation (LES) have generally been evaluated by their ability to predict single-time statistics of turbulent flows such as kinetic energy and Reynolds stresses. Recent application- of large eddy simulation to the evaluation of sound sources in turbulent flows, a problem in which time, correlations determine the frequency distribution of acoustic radiation, suggest that subgrid models should also be evaluated by their ability to predict time correlations in turbulent flows. This paper compares the two-point, two-time Eulerian velocity correlation evaluated from direct numerical simulation (DNS) with that evaluated from LES, using a spectral eddy viscosity, for isotropic homogeneous turbulence. It is found that the LES fields are too coherent, in the sense that their time correlations decay more slowly than the corresponding time. correlations in the DNS fields. This observation is confirmed by theoretical estimates of time correlations using the Taylor expansion technique. Tile reason for the slower decay is that the eddy viscosity does not include the random backscatter, which decorrelates fluid motion at large scales. An effective eddy viscosity associated with time correlations is formulated, to which the eddy viscosity associated with energy transfer is a leading order approximation.

He, Guowei↗