Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Statistical error”

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 1,405 records · Page 78
All sourcesNASA NTRSDOE OSTI
DOE OSTI2012

D-Factor: A Quantitative Model of Application Slow-Down in Multi-Resource Shared Systems

Scheduling multiple jobs onto a platform enhances system utilization by sharing resources. The benefits from higher resource utilization include reduced cost to construct, operate, and maintain a system, which often include energy consumption. Maximizing these benefits comes at a price - resource contention among jobs increases job completion time. In this paper, we analyze slow-downs of jobs due to contention for multiple resources in a system; referred to as dilation factor. We observe that multiple-resource contention creates non-linear dilation factors of jobs. From this observation, we establish a general quantitative model for dilation factors of jobs in multi-resource systems. A job is characterized by a vector-valued loading statistics and dilation factors of a job set are given by a quadratic function of their loading vectors. We demonstrate how to systematically characterize a job, maintain the data structure to calculate the dilation factor (loading matrix), and calculate the dilation factor of each job. We validate the accuracy of the model with multiple processes running on a native Linux server, virtualized servers, and with multiple MapReduce workloads co-scheduled in a cluster. Evaluation with measured data shows that the D-factor model has an error margin of less than 16%. We also show that the model can be integrated with an existing on-line scheduler to minimize the makespan of workloads.

99 GENERAL AND MISCELLANEOUS↗
DOE OSTI2015

Physics-based statistical learning approach to mesoscopic model selection

In materials science and many other research areas, models are frequently inferred without considering their generalization to unseen data. Here, we apply statistical learning using cross-validation to obtain an optimally predictive coarse-grained description of a two-dimensional kinetic nearest-neighbor Ising model with Glauber dynamics (GD) based on the stochastic Ginzburg-Landau equation (sGLE). The latter is learned from GD “training” data using a log-likelihood analysis, and its predictive ability for various complexities of the model is tested on GD “test” data independent of the data used to train the model on. Using two different error metrics, we perform a detailed analysis of the error between magnetization time trajectories simulated using the learned sGLE coarse-grained description and those obtained using the GD model. We show that both for equilibrium and out-of-equilibrium GD training trajectories, the standard phenomenological description using a quartic free energy does not always yield the most predictive coarse-grained model. Moreover, increasing the amount of training data can shift the optimal model complexity to higher values. Our results are promising in that they pave the way for the use of statistical learning as a general tool for materials modeling and discovery.

36 MATERIALS SCIENCE↗
DOE OSTI2014

Data free inference with processed data products

Here, we consider the context of probabilistic inference of model parameters given error bars or confidence intervals on model output values, when the data is unavailable. We introduce a class of algorithms in a Bayesian framework, relying on maximum entropy arguments and approximate Bayesian computation methods, to generate consistent data with the given summary statistics. Once we obtain consistent data sets, we pool the respective posteriors, to arrive at a single, averaged density on the parameters. This approach allows us to perform accurate forward uncertainty propagation consistent with the reported statistics.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2011

Inter- and Intrafraction Patient Positioning Uncertainties for Intracranial Radiotherapy: A Study of Four Frameless, Thermoplastic Mask-Based Immobilization Strategies Using Daily Cone-Beam CT

Purpose: To determine whether frameless thermoplastic mask-based immobilization is adequate for image-guided cranial radiosurgery. Methods and Materials: Cone-beam CT localization data from patients with intracranial tumors were studied using daily pre- and posttreatment scans. The systems studied were (1) Type-S IMRT (head only) mask (Civco) with head cushion; (2) Uni-Frame mask (Civco) with head cushion, coupled with a BlueBag body immobilizer (Medical Intelligence); (3) Type-S head and shoulder mask with head and shoulder cushion (Civco); (4) same as previous, coupled with a mouthpiece. The comparative metrics were translational shift magnitude and average rotation angle; systematic inter-, random inter-, and random intrafraction positioning error was computed. For strategies 1-4, respectively, the analysis for interfraction variability included data from 20, 9, 81, and 11 patients, whereas that for intrafraction variability included a subset of 7, 9, 16, and 8 patients. The results were compared for statistical significance using an analysis of variance test. Results: Immobilization system 4 provided the best overall accuracy and stability. The mean interfraction translational shifts ({+-} SD) were 2.3 ({+-} 1.4), 2.2 ({+-} 1.1), 2.7 ({+-} 1.5), and 2.1 ({+-} 1.0) mm whereas intrafraction motion was 1.1 ({+-} 1.2), 1.1 ({+-} 1.1), 0.7 ({+-} 0.9), and 0.7 ({+-} 0.8) mm for devices 1-4, respectively. No significant correlation between intrafraction motion and treatment time was evident, although intrafraction motion was not purely random. Conclusions: We find that all frameless thermoplastic mask systems studied are viable solutions for image-guided intracranial radiosurgery. With daily pretreatment corrections, symmetric PTV margins of 1 mm would likely be adequate if ideal radiation planning and targeting systems were available.

62 RADIOLOGY AND NUCLEAR MEDICINE↗
DOE OSTI2014

Ab initio molecular dynamics with noisy forces: Validating the quantum Monte Carlo approach with benchmark calculations of molecular vibrational properties

We present a systematic study of a recently developed ab initio simulation scheme based on molecular dynamics and quantum Monte Carlo. In this approach, a damped Langevin molecular dynamics is employed by using a statistical evaluation of the forces acting on each atom by means of quantum Monte Carlo. This allows the use of an highly correlated wave function parametrized by several variational parameters and describing quite accurately the Born-Oppenheimer energy surface, as long as these parameters are determined at the minimum energy condition. However, in a statistical method both the minimization method and the evaluation of the atomic forces are affected by the statistical noise. In this work, we study systematically the accuracy and reliability of this scheme by targeting the vibrational frequencies of simple molecules such as the water monomer, hydrogen sulfide, sulfur dioxide, ammonia, and phosphine. We show that all sources of systematic errors can be controlled and reliable frequencies can be obtained with a reasonable computational effort. This work provides convincing evidence that this molecular dynamics scheme can be safely applied also to realistic systems containing several atoms.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2019

Inferring Atmospheric Release Characteristics in a Large Computer Experiment using Bayesian Adaptive Splines

An atmospheric release of hazardous material, whether accidental or intentional, can be catastrophic for those in the path of the plume. Predicting the path of a plume based on characteristics of the release (location, amount and duration) and meteorological conditions is an active research area highly relevant for emergency and long-term response to these releases. As a result, researchers have developed particle dispersion simulators to provide plume path predictions that incorporate release characteristics and meteorological conditions. However, since release characteristics and meteorological conditions are often unknown, the inverse problem is of great interest, that is, based on all the observations of the plume so far, what can be inferred about the release characteristics? This is the question we seek to answer using plume observations from a controlled release at the Diablo Canyon Nuclear Power Plant in Central California. With access to a large number of evaluations of a computationally expensive particle dispersion simulator that includes continuous and categorical inputs and spatio-temporal output, building a fast statistical surrogate model (or emulator) presents many statistical challenges, but is an essential tool for inverse modeling and sensitivity analysis. We achieve accurate emulation using Bayesian adaptive splines to model weights on empirical orthogonal functions. Here, we use this emulator as well as appropriately identifiable simulator discrepancy and observational error models to calibrate the simulator, thus finding a posterior distribution of characteristics of the release. Since the release was controlled, these characteristics are known, making it possible to compare our findings to the truth.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2015

Degradation Mechanisms and Lifetime Prediction for Lithium-Ion Batteries -- A Control Perspective: Preprint

Predictive models of Li-ion battery lifetime must consider a multiplicity of electrochemical, thermal, and mechanical degradation modes experienced by batteries in application environments. To complicate matters, Li-ion batteries can experience different degradation trajectories that depend on storage and cycling history of the application environment. Rates of degradation are controlled by factors such as temperature history, electrochemical operating window, and charge/discharge rate. We present a generalized battery life prognostic model framework for battery systems design and control. The model framework consists of trial functions that are statistically regressed to Li-ion cell life datasets wherein the cells have been aged under different levels of stress. Degradation mechanisms and rate laws dependent on temperature, storage, and cycling condition are regressed to the data, with multiple model hypotheses evaluated and the best model down-selected based on statistics. The resulting life prognostic model, implemented in state variable form, is extensible to arbitrary real-world scenarios. The model is applicable in real-time control algorithms to maximize battery life and performance. We discuss efforts to reduce lifetime prediction error and accommodate its inevitable impact in controller design.

25 ENERGY STORAGE↗
DOE OSTI2016

A validated tumor control probability model based on a meta-analysis of low, intermediate, and high-risk prostate cancer patients treated by photon, proton, or carbon-ion radiotherapy

Purpose: A fully heterogeneous population averaged mechanistic tumor control probability (TCP) model is appropriate for the analysis of external beam radiotherapy (EBRT). This has been accomplished for EBRT photon treatment of intermediate-risk prostate cancer. Extending the TCP model for low and high-risk patients would be beneficial in terms of overall decision making. Furthermore, different radiation treatment modalities such as protons and carbon-ions are becoming increasingly available. Consequently, there is a need for a complete TCP model. Methods: A TCP model was fitted and validated to a primary endpoint of 5-year biological no evidence of disease clinical outcome data obtained from a review of the literature for low, intermediate, and high-risk prostate cancer patients (5218 patients fitted, 1088 patients validated), treated by photons, protons, or carbon-ions. The review followed the preferred reporting item for systematic reviews and meta-analyses statement. Treatment regimens include standard fractionation and hypofractionation treatments. Residual analysis and goodness of fit statistics were applied. Results: The TCP model achieves a good level of fit overall, linear regression results in a p-value of <0.000 01 with an adjusted-weighted-R{sup 2} value of 0.77 and a weighted root mean squared error (wRMSE) of 1.2%, to the fitted clinical outcome data. Validation of the model utilizing three independent datasets obtained from the literature resulted in an adjusted-weighted-R{sup 2} value of 0.78 and a wRMSE of less than 1.8%, to the validation clinical outcome data. The weighted mean absolute residual across the entire dataset is found to be 5.4%. Conclusions: This TCP model fitted and validated to clinical outcome data, appears to be an appropriate model for the inclusion of all clinical prostate cancer risk categories, and allows evaluation of current EBRT modalities with regard to tumor control prediction.

61 RADIATION PROTECTION AND DOSIMETRY↗
DOE OSTI2019

Cluster Cosmology with the Velocity Distribution Function of the HeCS-SZ Sample

We apply the velocity distribution function (VDF) to a sample of Sunyaev–Zel'dovich (SZ)-selected clusters, and we report preliminary cosmological constraints in the ${\sigma }_{8}$-${{\rm{\Omega }}}_{m}$ cosmological parameter space. The VDF is a forward-modeled test statistic that can be used to constrain cosmological models directly from galaxy cluster dynamical observations. The method was introduced in Ntampaka et al. and employs line-of-sight velocity measurements to directly constrain cosmological parameters; it is less sensitive to measurement error than a standard halo mass function approach. The method is applied to the Hectospec Survey of Sunyaev–Zeldovich-Selected Clusters sample, which is a spectroscopic follow-up of a Planck-selected sample of 83 galaxy clusters. Credible regions are calculated by comparing the VDF of the observed cluster sample to that of mock observations, yielding ${{ \mathcal S }}_{8}$ $\equiv \,{\sigma }_{8}{\left({{\rm{\Omega }}}_{m}/0.3\right)}^{0.25}=0.751\pm 0.037$. These constraints are in tension with the Planck Cosmic Microwave Background TT fiducial value, which lies outside of our 95% credible region, but are in agreement with some recent analyses of large-scale structure that observe fewer massive clusters than are predicted by the Planck fiducial cosmological parameters.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2012

Sub-daily Statistical Downscaling of Meteorological Variables Using Neural Networks

A new open source neural network temporal downscaling model is described and tested using CRU-NCEP reanal ysis and CCSM3 climate model output. We downscaled multiple meteorological variables in tandem from monthly to sub-daily time steps while also retaining consistent correlations between variables. We found that our feed forward, error backpropagation approach produced synthetic 6 hourly meteorology with biases no greater than 0.6% across all variables and variance that was accurate within 1% for all variables except atmospheric pressure, wind speed, and precipitation. Correlations between downscaled output and the expected (original) monthly means exceeded 0.99 for all variables, which indicates that this approach would work well for generating atmospheric forcing data consistent with mass and energy conserved GCM output. Our neural network approach performed well for variables that had correlations to other variables of about 0.3 and better and its skill was increased by downscaling multiple correlated variables together. Poor replication of precipitation intensity however required further post-processing in order to obtain the expected probability distribution. The concurrence of precipitation events with expected changes in sub ordinate variables (e.g., less incident shortwave radiation during precipitation events) were nearly as consistent in the downscaled data as in the training data with probabilities that differed by no more than 6%. Our downscaling approach requires training data at the target time step and relies on a weak assumption that climate variability in the extrapolated data is similar to variability in the training data.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2015

Constraints on gravity and dark energy from the pairwise kinematic Sunyaev–Zel’Dovich effect

We calculate the constraints on dark energy and cosmic modifications to gravity achievable with upcoming cosmic microwave background (CMB) surveys sensitive to the Sunyaev–Zel’dovich (SZ) effects. The analysis focuses on using the mean pairwise velocity of clusters as observed through the kinematic SZ effect (kSZ), an approach based on the same methods used for the first detection of the kSZ effect, and includes a detailed derivation and discussion of this statistic’s covariance under a variety of different survey assumptions. The potential of current, Stage II, and upcoming, Stages III and IV, CMB observations are considered, in combination with contemporaneous spectroscopic and photometric galaxy observations. A detailed assessment is made of the sensitivity to the assumed statistical and systematic uncertainties in the optical depth determination, the magnitude and uncertainty in the minimum detectable mass, and the importance of pairwise velocity correlations at small separations, where nonlinear effects can start to arise. In combination with Stage III constraints on the expansion history, such as those projected by the Dark Energy Task Force, we forecast 5% and 3% for fractional errors on the growth factor, γ, for Stage III and IV surveys, respectively, and 2% constraints on the growth rate, f{sub g}, for a Stage IV survey for 0.2<z<0.6. The results suggest that kSZ measurements of cluster peculiar velocities, obtained from cross-correlation with upcoming spectroscopic galaxy surveys, could provide robust tests of dark energy and theories of gravity on cosmic scales.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2016

Quantifying element incorporation in multispecies biofilms using nanoscale secondary ion mass spectrometry image analysis

Elucidating nutrient exchange in microbial communities is an important step in understanding the relationships between microbial systems and global biogeochemical cycles, but these communities are complex and the interspecies interactions that occur within them are not well understood. Phototrophic consortia are useful and relevant experimental systems to investigate such interactions as they are not only prevalent in the environment, but some are cultivable in vitro and amenable to controlled scientific experimentation. Nanoscale secondary ion mass spectrometry (NanoSIMS) is a powerful, high spatial resolution tool capable of visualizing the metabolic activities of single cells within a biofilm, but quantitative analysis of the resulting data has typically been a manual process, resulting in a task that is both laborious and susceptible to human error. In this work, the authors describe the creation and application of a semiautomated image-processing pipeline that can analyze NanoSIMS-generated data, applied to phototrophic biofilms as an example. The tool employs an image analysis process, which includes both elemental and morphological segmentation, producing a final segmented image that allows for discrimination between autotrophic and heterotrophic biomass, the detection of individual cyanobacterial filaments and heterotrophic cells, the quantification of isotopic incorporation of individual heterotrophic cells, and calculation of relevant population statistics. The authors demonstrate the functionality of the tool by using it to analyze the uptake of 15 N provided as either nitrate or ammonium through the unicyanobacterial consortium UCC-O and imaged via NanoSIMS. The authors found that the degree of 15N incorporation by individual cells was highly variable when labeled with 15 NH 4 + , but much more even when biofilms were labeled with 15 NO 3 – . In the 15 NH 4 + -amended biofilms, the heterotrophic distribution of 15 N incorporation was highly skewed, with a large population showing moderate 15 N incorporation and a small number of organisms displaying very high 15 N uptake. The results showed that analysis of NanoSIMS data can be performed in a way that allows for quantitation of the elemental uptake of individual cells, a technique necessary for advancing research into the metabolic networks that exist within biofilms with statistical analyses that are supported by automated, user-friendly processes.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2012

Bayesian Neural Networks for Uncertainty Analysis of Hydrologic Modeling: A Comparison of Two Schemes

Bayesian Neural Networks (BNNs) have been shown as useful tools to analyze modeling uncertainty of Neural Networks (NNs). This research focuses on the comparison of two BNNs. The first BNNs (BNN-I) use statistical methods to describe the characteristics of different uncertainty sources (input, parameter, and model structure) and integrate these uncertainties into a Markov Chain Monte Carlo (MCMC) framework to estimate total uncertainty. The second BNNs (BNN-II) lump all uncertainties into a single error term (i.e. the residual between model prediction and measurement). In this study, we propose a simple BNN-II, which use Genetic Algorithms (GA) and Bayesian Model Averaging (BMA) to calibrate Neural Networks with different structures (number of hidden units) and combine the predictions from different NNs to derive predictions and uncertainty analysis. We tested these two BNNs in two watersheds for daily and monthly hydrologic simulation. The BMA based BNNs developed in this study outperforms BNN-I in the two watersheds in terms of both accurate prediction and uncertainty estimation. These results show that, given incomplete understanding of the characteristics associated with each uncertainty source, the simple lumped error approach may yield better prediction and uncertainty estimation.

99 GENERAL AND MISCELLANEOUS↗
DOE OSTI2015

MODELING THE ANOMALY OF SURFACE NUMBER DENSITIES OF GALAXIES ON THE GALACTIC EXTINCTION MAP DUE TO THEIR FIR EMISSION CONTAMINATION

The most widely used Galactic extinction map is constructed assuming that the observed far-infrared (FIR) fluxes come entirely from Galactic dust. According to the earlier suggestion by Yahata et al., we consider how FIR emission of galaxies affects the SFD map. We first compute the surface number density of Sloan Digital Sky Survey (SDSS) DR7 galaxies as a function of the r-band extinction, A {sub r,} {sub SFD}. We confirm that the surface densities of those galaxies positively correlate with A {sub r,} {sub SFD} for A {sub r,} {sub SFD} < 0.1, as first discovered by Yahata et al. for SDSS DR4 galaxies. Next we construct an analytical model to compute the surface density of galaxies, taking into account the contamination of their FIR emission. We adopt a log-normal probability distribution for the ratio of 100 μm and r-band luminosities of each galaxy, y ≡ (νL){sub 100} {sub μm}/(νL) {sub r}. Then we search for the mean and rms values of y that fit the observed anomaly, using the analytical model. The required values to reproduce the anomaly are roughly consistent with those measured from the stacking analysis of SDSS galaxies. Due to the limitation of our statistical modeling, we are not yet able to remove the FIR contamination of galaxies from the extinction map. Nevertheless, the agreement with the model prediction suggests that the FIR emission of galaxies is mainly responsible for the observed anomaly. Whereas the corresponding systematic error in the Galactic extinction map is 0.1-1 mmag, it is directly correlated with galaxy clustering and thus needs to be carefully examined in precision cosmology.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2015

A COUPLED 2 × 2D BABCOCK–LEIGHTON SOLAR DYNAMO MODEL. I. SURFACE MAGNETIC FLUX EVOLUTION

The need for reliable predictions of the solar activity cycle motivates the development of dynamo models incorporating a representation of surface processes sufficiently detailed to allow assimilation of magnetographic data. In this series of papers we present one such dynamo model, and document its behavior and properties. This first paper focuses on one of the model’s key components, namely surface magnetic flux evolution. Using a genetic algorithm, we obtain best-fit parameters of the transport model by least-squares minimization of the differences between the associated synthetic synoptic magnetogram and real magnetographic data for activity cycle 21. Our fitting procedure also returns Monte Carlo-like error estimates. We show that the range of acceptable surface meridional flow profiles is in good agreement with Doppler measurements, even though the latter are not used in the fitting process. Using a synthetic database of bipolar magnetic region (BMR) emergences reproducing the statistical properties of observed emergences, we also ascertain the sensitivity of global cycle properties, such as the strength of the dipole moment and timing of polarity reversal, to distinct realizations of BMR emergence, and on this basis argue that this stochasticity represents a primary source of uncertainty for predicting solar cycle characteristics.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2010

FAST AND OPTIMAL COSMIC MICROWAVE BACKGROUND LENSING USING STATISTICAL INTERPOLATION ON THE SPHERE

We describe an accurate and fast pixel-based statistical method to interpolate fields of arbitrary spin on the sphere. We call this method Fast and Lean Interpolation on the Sphere (FLINTS). The method predicts the optimal interpolated values based on the theory of isotropic Gaussian random fields and provides an accurate error estimate at no additional cost. We use this method to compute lensed cosmic microwave background (CMB) maps precisely and quickly, achieving a relative precision of 0.02% at a HEALPIX resolution of N {sub side} = 4096, for a bandlimit of l{sub max}= 4096 in the same time it takes to simulate the original, unlensed CMB map. The method is suitable for efficient, distributed memory parallelization. The power spectra of our lensed maps are accurate to better than 0.5% at l= 3000 for the temperature, the E and B modes of the polarization. As expected theoretically, we demonstrate that, on realistic cases, this method is between two and three orders of magnitude more precise than other known interpolation methods for the same computational cost.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2011

KEPLER-10 c: A 2.2 EARTH RADIUS TRANSITING PLANET IN A MULTIPLE SYSTEM

The Kepler mission has recently announced the discovery of Kepler-10 b, the smallest exoplanet discovered to date and the first rocky planet found by the spacecraft. A second, 45 day period transit-like signal present in the photometry from the first eight months of data could not be confirmed as being caused by a planet at the time of that announcement. Here we apply the light curve modeling technique known as BLENDER to explore the possibility that the signal might be due to an astrophysical false positive (blend). To aid in this analysis we report the observation of two transits with the Spitzer Space Telescope at 4.5 {mu}m. When combined, they yield a transit depth of 344 {+-} 85 ppm that is consistent with the depth in the Kepler passband (376 {+-} 9 ppm, ignoring limb darkening), which rules out blends with an eclipsing binary of a significantly different color than the target. Using these observations along with other constraints from high-resolution imaging and spectroscopy, we are able to exclude the vast majority of possible false positives. We assess the likelihood of the remaining blends, and arrive conservatively at a false alarm rate of 1.6 x 10{sup -5} that is small enough to validate the candidate as a planet (designated Kepler-10 c) with a very high level of confidence. The radius of this object is measured to be R{sub p} = 2.227{sup +0.052}{sub -0.057} R{sub +} (in which the error includes the uncertainty in the stellar properties), but currently available radial-velocity measurements only place an upper limit on its mass of about 20 M{sub +}. Kepler-10 c represents another example (with Kepler-9 d and Kepler-11 g) of statistical 'validation' of a transiting exoplanet, as opposed to the usual 'confirmation' that can take place when the Doppler signal is detected or transit timing variations are measured. It is anticipated that many of Kepler's smaller candidates will receive a similar treatment since dynamical confirmation may be difficult or impractical with the sensitivity of current instrumentation.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2025

Describing Point Defect Topology in 2D Energy Materials through Computer Vision

Point defects such as vacancies and impurity atoms strongly impact the performance of 2D materials. Traditional efforts often rely on manual detection, a process that is time-intensive, prone to human error, and challenging to scale. Here we leverage machine learning (ML) methods to identify and quantify vacancies within 2D transition metal carbides (Ti3C2, MXenes), aiming to expedite detection while improving accuracy. MXenes exhibit valuable defect-defined electrochemical properties, but we currently lack statistical understanding of defect topology needed to fully harness these materials. Here we employ a convolutional neural network for semantic segmentation of experimental MXene images, opening an opportunity to conduct a rigorous statistical study on defect hierarchy while investigating local relaxation in the lattice. We show how the integration of ML can yield fundamental insight into point defects, providing a powerful tool that will play an increasingly crucial role in the future of materials science.

2d materials↗