Search NASA⌕ Search

SEARCH · Search NASA

Results for “predict”

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 703 records · Page 39

Direct Prediction of Phonon Density of States With Euclidean Neural Networks

Abstract Machine learning has demonstrated great power in materials design, discovery, and property prediction. However, despite the success of machine learning in predicting discrete properties, challenges remain for continuous property prediction. The challenge is aggravated in crystalline solids due to crystallographic symmetry considerations and data scarcity. Here, the direct prediction of phonon density‐of‐states (DOS) is demonstrated using only atomic species and positions as input. Euclidean neural networks are applied, which by construction are equivariant to 3D rotations, translations, and inversion and thereby capture full crystal symmetry, and achieve high‐quality prediction using a small training set of examples with over 64 atom types. The predictive model reproduces key features of experimental data and even generalizes to materials with unseen elements, and is naturally suited to efficiently predict alloy systems without additional computational cost. The potential of the network is demonstrated by predicting a broad number of high phononic specific heat capacity materials. The work indicates an efficient approach to explore materials' phonon structure, and can further enable rapid screening for high‐performance thermal storage materials and phonon‐mediated superconductors.

97 MATHEMATICS AND COMPUTING↗

Protein location prediction using atomic composition and global features of the amino acid sequence

Subcellular location of protein is constructive information in determining its function, screening for drug candidates, vaccine design, annotation of gene products and in selecting relevant proteins for further studies. Computational prediction of subcellular localization deals with predicting the location of a protein from its amino acid sequence. For a computational localization prediction method to be more accurate, it should exploit all possible relevant biological features that contribute to the subcellular localization. In this work, we extracted the biological features from the full length protein sequence to incorporate more biological information. A new biological feature, distribution of atomic composition is effectively used with, multiple physiochemical properties, amino acid composition, three part amino acid composition, and sequence similarity for predicting the subcellular location of the protein. Support Vector Machines are designed for four modules and prediction is made by a weighted voting system. Our system makes prediction with an accuracy of 100, 82.47, 88.81 for self-consistency test, jackknife test and independent data test respectively. Our results provide evidence that the prediction based on the biological features derived from the full length amino acid sequence gives better accuracy than those derived from N-terminal alone. Considering the features as a distribution within the entire sequence will bring out underlying property distribution to a greater detail to enhance the prediction accuracy.

60 APPLIED LIFE SCIENCES↗

Predicting concrete compressive strength using hybrid ensembling of surrogate machine learning models

This study aims to implement a hybrid ensemble surrogate machine learning technique in predicting the compressive strength (CS) of concrete, an important parameter used for durability design and service life prediction of concrete structures in civil engineering projects. For this purpose, an experimental database consisting of 1030 records has been compiled from the machine learning repository of the University of California, Irvine. The database was used to train and validate four conventional machine learning (CML) models, namely Artificial Neural Network (ANN), Linear and Non-Linear Multivariate Adaptive Regression Splines (MARS-L and MARS-C), Gaussian Process Regression (GPR), and Minimax Probability Machine Regression (MPMR). Subsequently, the predicted outputs of CML models were combined and trained using ANN to construct the Hybrid Ensemble Model (HENSM). It is observed that the proposed HENSM produces higher predictive accuracy compared to the CML models used in the present study. The predictive performance of all models for CS prediction was compared using the testing dataset and it is found that the HENSM model attained the highest predictive accuracy in both phases. Based on the experimental results, the newly constructed HENSM model is very potential to be a new alternative in handling the overfitting issues of CML models and hence, can be used to predict the concrete CS, including the design of less polluting and more sustainable concrete constructions.

36 MATERIALS SCIENCE↗

Incorporating machine learning with building network analysis to predict multi-building energy use

Predicting multi-building energy use at campus or city district scale has recently gained more attention; and more researchers have started to define reference buildings and study inter-impact between building groups. However, how to integrate the relationship to define reference buildings and predict multi-building energy use, using significantly less amount of building data and reducing complexity of prediction models, remains an open research question. To resolve this, this study proposed a novel method to predict multi-building energy use by integrating a social network analysis (SNA) with an Artificial Neural Network (ANN) technique. The SNA method was used to establish a building network (BN) by identifying reference buildings and determine correlations between reference buildings and non-reference buildings. The ANN technique was applied to learn correlations and historical building energy use, and then used to predict multi-building energy use. To validate the SNA-ANN method, 17 buildings in the Southeast University campus, located in Nanjing, China, were studied. These buildings have three years of actual monthly electricity use data and were grouped into four types: office, educational, laboratory, and residential. The results showed the integrated SNA-ANN method achieved average prediction accuracies of 90.67% for the office group, 90.79% for the educational group, 92.34% for the laboratory group, and 83.32% for the residential group. The results demonstrated the proposed SNA-ANN method achieved an accuracy of 90.28% for the predicted energy use for all building groups. Finally, this study provides insights into advancing the interdisciplinary research on multi-building energy use prediction.

29 ENERGY PLANNING, POLICY, AND ECONOMY↗

Predicting the proximity to macroscopic failure using local strain populations from dynamic in situ X-ray tomography triaxial compression experiments on rocks

Predicting the proximity of large-scale dynamic failure is a critical concern in the engineering and geophysical sciences. Here we use evolving contractive, dilatational, and shear strain deformation preceding failure in dynamic X-ray tomography experiments to examine which strain components best predict the proximity to failure. We develop machine learning models to predict the proximity to failure using time series of three-dimensional local incremental strain tensor fields acquired in rock deformation experiments under stress conditions of the upper crust. Three-dimensional scans acquired in situ throughout triaxial compression experiments provide a distribution of density contrasts from which we estimate the three-dimensional incremental strain that accumulates between each scan acquisition. Training machine learning models on multiple experiments of six rock types provides suites of feature importance that indicate the predictive power of each feature. Comparing the average importance of groups of features that include information about each strain component quantifies the ability of the contractive, dilatational and shear strain to predict the proximity of macroscopic failure. A total of 24 models of four machine learning algorithms with six rock types indicate that 1) the dilatational strain provides the best predictive power of the strain components, and 2) the intermediate values (25th-75th percentile) of the strain population provide the best predictive power of the statistics of the strain populations. In addition, the success of the predictions of models trained on one rock type and tested on other rock types quantifies the similarities and differences of the precursory strain accumulation process in the six rock types. These similarities suggest the potential existence of a unified theory of brittle rock deformation for a range of rock types.

58 GEOSCIENCES↗

Prediction of Response to Preoperative Chemoradiotherapy in Rectal Cancer by Multiplex Kinase Activity Profiling

Purpose: Tumor response of rectal cancer to preoperative chemoradiotherapy (CRT) varies considerably. In experimental tumor models and clinical radiotherapy, activity of particular subsets of kinase signaling pathways seems to predict radiation response. This study aimed to determine whether tumor kinase activity profiles might predict tumor response to preoperative CRT in locally advanced rectal cancer (LARC). Methods and Materials: Sixty-seven LARC patients were treated with a CRT regimen consisting of radiotherapy, fluorouracil, and, where possible, oxaliplatin. Pretreatment tumor biopsy specimens were analyzed using microarrays with kinase substrates, and the resulting substrate phosphorylation patterns were correlated with tumor response to preoperative treatment as assessed by histomorphologic tumor regression grade (TRG). A predictive model for TRG scores from phosphosubstrate signatures was obtained by partial-least-squares discriminant analysis. Prediction performance was evaluated by leave-one-out cross-validation and use of an independent test set. Results: In the patient population, 73% and 15% were scored as good responders (TRG 1-2) or intermediate responders (TRG 3), whereas 12% were assessed as poor responders (TRG 4-5). In a subset of 7 poor responders and 12 good responders, treatment outcome was correctly predicted for 95%. Application of the prediction model on the remaining patient samples resulted in correct prediction for 85%. Phosphosubstrate signatures generated by poor-responding tumors indicated high kinase activity, which was inhibited by the kinase inhibitor sunitinib, and several discriminating phosphosubstrates represented proteins derived from signaling pathways implicated in radioresistance. Conclusions: Multiplex kinase activity profiling may identify functional biomarkers predictive of tumor response to preoperative CRT in LARC.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Assessing the capability of continuum and discrete particle methods to simulate gas-solids flow using DNS predictions as a benchmark

For this study, gas–solids flow in a three-dimension periodic domain was numerically investigated by direct numerical simulation (DNS), computational fluid dynamic-discrete element method (CFD-DEM) and two-fluid model (TFM). DNS data obtained by finely resolving the flow around every particle are used as a benchmark to assess the validity of coarser DEM and TFM approaches. The CFD-DEM predicts the correct cluster size distribution and under-predicts the macro-scale slip velocity even with a grid size as small as twice the particle diameter. The TFM approach predicts larger cluster size and lower slip velocity with a homogeneous drag correlation. Although the slip velocity can be matched by a simple modification to the drag model, the predicted voidage distribution is still different from DNS: Both CFD-DEM and TFM over-predict the fraction of particles in dense regions and under-predict the fraction of particles in regions of intermediate void fractions. Also, the cluster aspect ratio of DNS is smaller than CFD-DEM and TFM. Since a simple correction to the drag model can predict a correct slip velocity, it is hopeful that drag corrections based on more elaborate theories that consider voidage gradient and particle fluctuations may be able to improve the current predictions of cluster distribution.

42 ENGINEERING↗

Predicting human developmental toxicity of pharmaceuticals using human embryonic stem cells and metabolomics

Teratogens, substances that may cause fetal abnormalities during development, are responsible for a significant number of birth defects. Animal models used to predict teratogenicity often do not faithfully correlate to human response. Here, we seek to develop a more predictive developmental toxicity model based on an in vitro method that utilizes both human embryonic stem (hES) cells and metabolomics to discover biomarkers of developmental toxicity. We developed a method where hES cells were dosed with several drugs of known teratogenicity then LC-MS analysis was performed to measure changes in abundance levels of small molecules in response to drug dosing. Statistical analysis was employed to select for specific mass features that can provide a prediction of the developmental toxicity of a substance. These molecules can serve as biomarkers of developmental toxicity, leading to better prediction of teratogenicity. In particular, our work shows a correlation between teratogenicity and changes of greater than 10% in the ratio of arginine to asymmetric dimethylarginine levels. In addition, this study resulted in the establishment of a predictive model based on the most informative mass features. This model was subsequently tested for its predictive accuracy in two blinded studies using eight drugs of known teratogenicity, where it correctly predicted the teratogenicity for seven of the eight drugs. Thus, our initial data shows that this platform is a robust alternative to animal and other in vitro models for the prediction of the developmental toxicity of chemicals that may also provide invaluable information about the underlying biochemical pathways.

60 APPLIED LIFE SCIENCES↗

PySIDT: Subgraph Isomorphic Decision Trees for Molecular Property Prediction

Accurate molecular property prediction is important across all fields of chemistry. Deep neural networks (DNNs) have become increasingly popular due to their ability to train automatically, avoiding the incredibly tedious process of constructing and extending traditional property estimation schemes. However, DNNs require large amounts of training data, are challenging to interpret, require large amounts of memory to load even during inference, and have severe difficulties incorporating qualitative chemical knowledge, which are often desired for molecular property prediction tasks. Here, in this study, we present PySIDT (https://github.com/zadorlab/PySIDT), a software for training and running inference on Subgraph Isomorphic Decision Trees (SIDTs). SIDTs are graph-based decision trees made of nodes associated with molecular substructures. Inference is done by descending target molecular structures down the decision tree to nodes with matching subgraph isomorphic substructures and making predictions based on the final (most specific) nodes matched. SIDTs scale down well to dataset sizes much smaller than is feasible for DNNs. As trees of molecular substructures, SIDTs are inherently readable and easy to visualize, making them easy to analyze. They are also straightforward to extend and retrain, facilitate uncertainty estimation, and enable easy integration of expert knowledge. We demonstrate the SIDT approach discussing its application to a diverse range of molecular prediction tasks: rate coefficient estimation, diffusion coefficient estimation, thermochemistry estimation, transition state bond stretch prediction, p K a prediction, stability of molecular structures, stability of surface structures, and prediction of surface lateral interaction energetics. Additionally, we demonstrate the power of the SIDT algorithms in two direct learning curve vanilla comparisons with the popular DNN-based software Chemprop and the popular gradient boosted trees-based software XGBoost on enthalpy of formation and rate coefficient prediction tasks. In particular, in the enthalpy of formation case, vanilla PySIDT is able to outperform vanilla Chemprop and XGBoost across the full range of training/validation set sizes out to 11,560 data points.

Johnson, Matthew Sean [Sandia National Laboratorie↗

Simplified Models for Accelerated Structural Prediction of Conjugated Semiconducting Polymers

We perform molecular dynamics simulations of poly(benzodithiophene-thienopyrrolodione) (BDT-TPD) oligomers in order to evaluate the accuracy with which unoptimized molecular models can predict experimentally characterized morphologies. The predicted morphologies are characterized using simulated grazing-incidence X-ray scattering (GIXS) and compared to the experimental scattering patterns. We find that approximating the aromatic rings in BDT-TPD with rigid bodies, rather than combinations of bond, angle, and dihedral constraints, results in 14% lower computational cost and provides nearly equivalent structural predictions compared to the flexible model case. The predicted glass transition temperature of BDT-TPD (410 +/- 32 K) is found to be in agreement with experiments. Predicted morphologies demonstrate short-range structural order due to stacking of the chain backbones (p-p stacking around 3.9 A), and long-range spatial correlations due to the self-organization of backbone stacks into 'ribbons' (lamellar ordering around 20.9 A), representing the best-to-date computational predictions of structure of complex conjugated oligomers. We find that expensive simulated annealing schedules are not needed to predict experimental structures here, with instantaneous quenches providing nearly equivalent predictions at a fraction of the computational cost of annealing. We therefore suggest utilizing rigid bodies and fast cooling schedules for high-throughput screening studies of semiflexible polymers and oligomers to utilize their significant computational benefits where appropriate.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Multimodel Predictive System for Carbon Dioxide Solubility in Saline Formation Waters

The prediction of carbon dioxide solubility in brine at conditions relevant to carbon sequestration (i.e., high temperature, pressure, and salt concentration (T-P-X)) is crucial when this technology is applied. Eleven mathematical models for predicting CO{sub 2} solubility in brine are compared and considered for inclusion in a multimodel predictive system. Model goodness of fit is evaluated over the temperature range 304–433 K, pressure range 74–500 bar, and salt concentration range 0–7 m (NaCl equivalent), using 173 published CO{sub 2} solubility measurements, particularly selected for those conditions. The performance of each model is assessed using various statistical methods, including the Akaike Information Criterion (AIC) and the Bayesian Information Criterion (BIC). Different models emerge as best fits for different subranges of the input conditions. A classification tree is generated using machine learning methods to predict the best-performing model under different T-P-X subranges, allowing development of a multimodel predictive system (MMoPS) that selects and applies the model expected to yield the most accurate CO{sub 2} solubility prediction. Statistical analysis of the MMoPS predictions, including a stratified 5-fold cross validation, shows that MMoPS outperforms each individual model and increases the overall accuracy of CO{sub 2} solubility prediction across the range of T-P-X conditions likely to be encountered in carbon sequestration applications.

54 ENVIRONMENTAL SCIENCES↗

Enhancing Skill of Initialized Decadal Predictions Using a Dynamic Model of Drift

Since near-term predictions of present-day climate are controlled by both initial condition predictability and boundary condition predictability, initialized prediction experiments aim to augment the external-forcing-related predictability realized in uninitialized projections with initial-condition-related predictability by appropriate observation-based initialization. However, and notwithstanding the considerable effort expended in finding such “good” initial states, a striking feature of current, state-of-the-art, initialized decadal hindcasts is their tendency to quickly drift away from the initialized state, with attendant loss of skill. We derive a dynamical model for such drift, and after validating it we show that including a recalibrated version of the model in a postprocessing framework is able to significantly augment the skill of initialized predictions beyond that achieved by a use of current techniques of postprocessing alone. We also show that the new methodology provides further crucial insights into issues related to initialized predictions.

58 GEOSCIENCES↗

A Bayesian Deep Learning Approach to Near-Term Climate Prediction

Since model bias and associated initialization shock are serious shortcomings that reduce prediction skills in state-of-the-art decadal climate prediction efforts, we pursue a complementary machine-learning-based approach to climate prediction. The example problem setting we consider consists of predicting natural variability of the North Atlantic sea surface temperature on the interannual timescale in the pre-industrial control simulation of the Community Earth System Model. While previous works have considered the use of recurrent networks such as convolutional LSTMs and reservoir computing networks in this and other similar problem settings, we currently focus on the use of feedforward convolutional networks. In particular, we find that a feedforward convolutional network with a Densenet architecture is able to outperform a convolutional LSTM in terms of predictive skill. Next, we go on to consider a probabilistic formulation of the same network based on Stein variational gradient descent and find that in addition to providing useful measures of predictive uncertainty, the probabilistic (Bayesian) version improves on its deterministic counterpart in terms of predictive skill. Finally, we characterize the reliability of the ensemble of machine learning models obtained in the probabilistic setting by using analysis tools developed in the context of ensemble numerical weather prediction.

54 ENVIRONMENTAL SCIENCES↗

Machine-learning informed prediction of high-entropy solid solution formation: Beyond the Hume-Rothery rules

The empirical rules for the prediction of solid solution formation proposed so far in the literature usually have very compromised predictability. Some rules with seemingly good predictability were, however, tested using small data sets. Based on an unprecedented large dataset containing 1252 multicomponent alloys, machine-learning methods showed that the formation of solid solutions can be very accurately predicted (93%). The machine-learning results help identify the most important features, such as molar volume, bulk modulus, and melting temperature. As such a new thermodynamics-based rule was developed to predict solid–solution alloys. The new rule is nonetheless slightly less accurate (73%) but has roots in the physical nature of the problem. The new rule is employed to predict solid solutions existing in the three blocks, each of which consists of 9 elements. The predictions encompass face-centered cubic (FCC), body-centered cubic (BCC), and hexagonal closest packed (HCP) structures in a high throughput manner. The validity of the prediction is further confirmed by CALculations of PHAse Diagram (CALPHAD) calculations with high consistency (94%). Since the new thermodynamics-based rule employs only elemental properties, applicability in screening for solid solution high-entropy alloys is straightforward and efficient.

36 MATERIALS SCIENCE↗

Soil microbiome predictability increases with spatial and taxonomic scale

Soil microorganisms shape ecosystem function, yet it remains an open question whether we can predict the composition of the soil microbiome in places before observing it. Furthermore, it is unclear whether the predictability of microbial life exhibits taxonomic- and spatial-scale dependence, as it does for macrobiological communities. Here, we leverage multiple large-scale soil microbiome surveys to develop predictive models of bacterial and fungal community composition in soil, then test these models against independent soil microbial community surveys from across the continental United States. We find remark- able scale dependence in community predictability. The predictability of bacterial and fungal communities increases with the spatial scale of observation, and fungal predictability increases with taxonomic scale. These patterns suggest that there is an increasing importance of deterministic versus stochastic processes with scale, consistent with findings in plant and animal communities, suggesting a general scaling relationship across biology. Biogeochemical functional groups and high-level taxonomic groups of microorganisms were equally predictable, indicating that traits and taxonomy are both powerful lenses for understanding soil communities. Here, by focusing on out-of-sample prediction, these findings suggest an emerging generality in our understanding of the soil microbiome, and that this understanding is fundamentally scale dependent

Biogeography↗

Incorrect computation of Madden-Julian oscillation prediction skill

The Madden–Julian oscillation (MJO) is a major tropical weather system and one of the largest sources of predictability for subseasonal-to-seasonal weather forecasts. Skillful prediction of the MJO has been a highly active area of research due to its large socio-economic impacts. Silini et al., herein S21, developed a machine learning model to predict the MJO, which they claimed to have an MJO prediction skill of 26–27 days over all seasons and 45 days for December–February (DJF) winter. If true, this would make the skill of their model competitive with that of the state-of-the-art dynamical MJO prediction systems at 20–35 days. However, here we show that the MJO prediction was calculated incorrectly in S21, which spuriously increased the performance of their model. Correctly computed skill of their model was substantially lower than that reported in S21; the skill for all seasons drops to 11–12 days and the skill for forecasts initialized during DJF drops to 15 days. Our findings clarify that the S21 machine learning model is not competitive with state-of-the-art numerical weather prediction models in predicting the MJO.

54 ENVIRONMENTAL SCIENCES↗

Machine learned synthesizability predictions aided by density functional theory

Abstract A grand challenge of materials science is predicting synthesis pathways for novel compounds. Data-driven approaches have made significant progress in predicting a compound’s synthesizability; however, some recent attempts ignore phase stability information. Here, we combine thermodynamic stability calculated using density functional theory with composition-based features to train a machine learning model that predicts a material’s synthesizability. Our model predicts the synthesizability of ternary 1:1:1 compositions in the half-Heusler structure, achieving a cross-validated precision of 0.82 and recall of 0.82. Our model shows improvement in predicting non-half-Heuslers compared to a previous study’s model, and identifies 121 synthesizable candidates out of 4141 unreported ternary compositions. More notably, 39 stable compositions are predicted unsynthesizable while 62 unstable compositions are predicted synthesizable; these findings otherwise cannot be made using density functional theory stability alone. This study presents a new approach for accurately predicting synthesizability, and identifies new half-Heuslers for experimental synthesis.

Lee, Andrew (ORCID:0000000153014295)↗

Advancing energy storage through solubility prediction: leveraging the potential of deep learning

Solubility prediction plays a crucial role in energy storage applications, such as redox flow batteries, because it directly affects the efficiency and reliability. Researchers have developed various methods that utilize quantum calculations and descriptors to predict the aqueous solubilities of organic molecules. Notably, machine learning models based on descriptors have shown promise for solubility prediction. As deep learning tools, graph neural networks (GNNs) have emerged to capture complex structure–property relationships for material property prediction. Specifically, MolGAT, a type of GNN model, was designed to incorporate n-dimensional edge attributes, enabling the modeling of intricacies in molecular graphs and enhancing the prediction capabilities. In a previous study, MolGAT successfully screened 23 467 promising redox-active molecules from a database of over 500 000 compounds, based on redox potential predictions. This study focused on applying the MolGAT model to predict the aqueous solubility (log S) of a broad range of organic compounds, including those previously screened for redox activity. The model was trained on a diverse sample of 8494 organic molecules from AqSolDB and benchmarked against literature data, demonstrating superior accuracy compared with other state of the art graph-based and descriptor-based models. Subsequently, the trained MolGAT model was employed to screen redox-active organic compounds identified in the first phase of high-throughput virtual screening, targeting favorable solubility in energy storage applications. The second round of screening, which considered solubility, yielded 12 332 promising redox-active and soluble organic molecules suitable for use in aqueous redox flow batteries. Thus, the two-phase high-throughput virtual screening approach utilizing MolGAT, specifically trained for redox potential and solubility, is an effective strategy for selecting suitable intrinsically soluble redox-active molecules from extensive databases, potentially advancing energy storage through reliable material development. This indicates that the model is reliable for predicting the solubility of various molecules and provides valuable insights for energy storage, pharmaceutical, environmental, and chemical applications.

25 ENERGY STORAGE↗