Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “machine learning, Random Forest”

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 217 records · Page 12
All sourcesNASA NTRSDOE OSTI
DOE OSTI2021

Classification of orthostatic intolerance through data analytics

Imbalance in the autonomic nervous system can lead to orthostatic intolerance manifested by dizziness, lightheadedness, and a sudden loss of consciousness (syncope); these are common conditions, but they are challenging to diagnose correctly. Uncertainties about the triggering mechanisms and the underlying pathophysiology have led to variations in their classification. This study uses machine learning to categorize patients with orthostatic intolerance. Here we use random forest classification trees to identify a small number of markers in blood pressure, and heart rate time-series data measured during head-up tilt to (a) distinguish patients with a single pathology and (b) examine data from patients with a mixed pathophysiology. Next, we use Kmeans to cluster the markers representing the time-series data. We apply the proposed method analyzing clinical data from 186 subjects identified as control or suffering from one of four conditions: postural orthostatic tachycardia (POTS), cardioinhibition, vasodepression, and mixed cardioinhibition and vasodepression. Classification results confirm the use of supervised machine learning. We were able to categorize more than 95% of patients with a single condition and were able to subgroup all patients with mixed cardioinhibitory and vasodepressor syncope. Clustering results confirm the disease groups and identify two distinct subgroups within the control and mixed groups. The proposed study demonstrates how to use machine learning to discover structure in blood pressure and heart rate time-series data. The methodology is used in classification of patients with orthostatic intolerance. Diagnosing orthostatic intolerance is challenging, and full characterization of the pathophysiological mechanisms remains a topic of ongoing research. This study provides a step toward leveraging machine learning to assist clinicians and researchers in addressing these challenges.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2022

Using soil library hyperspectral reflectance and machine learning to predict soil organic carbon: Assessing potential of airborne and spaceborne optical soil sensing

Soil organic carbon (SOC) is a key variable to determine soil functioning, ecosystem services, and global carbon cycles. Spectroscopy, particularly optical hyperspectral reflectance coupled with machine learning, can provide rapid, efficient, and cost-effective quantification of SOC. However, how to exploit soil hyperspectral reflectance to predict SOC concentration, and the potential performance of airborne and satellite data for predicting surface SOC at large scales remain relatively underknown. Here, this study utilized a continental-scale soil laboratory spectral library (37,540 full-pedon 350–2500 nm reflectance spectra with SOC concentration of 0–780 g·kg –1 across the US) to thoroughly evaluate seven machine learning algorithms including Partial-Least Squares Regression (PLSR), Random Forest (RF), K-Nearest Neighbors (KNN), Ridge, Artificial Neural Networks (ANN), Convolutional Neural Networks (CNN), and Long Short-Term Memory (LSTM) along with four preprocessed spectra, i.e. original, vector normalization, continuum removal, and first-order derivative, to quantify SOC concentration. Furthermore, by using the coupled soil-vegetation-atmosphere radiative transfer model, we simulated twelve airborne and spaceborne hyper/multi-spectral remote sensing data from surface bare soil laboratory spectra to evaluate their potential for estimating SOC concentration of surface bare soils. Results show that LSTM achieved best predictive performance of quantifying SOC concentration for the whole data sets (R 2 = 0.96, RMSE = 30.81 g·kg –1 ), mineral soils (SOC ≤ 120 g·kg –1 , R 2 = 0.71, RMSE = 10.60 g·kg –1 ), and organic soils (SOC > 120 g·kg –1 , R 2 = 0.78, RMSE = 62.31 g·kg –1 ). Spectral data preprocessing, particularly the first-order derivative, improved the performance of PLSR, RF, Ridge, KNN, and ANN, but not LSTM or CNN. We found that the SOC models of mineral and organic soils should be distinguished given their distinct spectral signatures. Finally, we identified that the shortwave infrared is vital for airborne and spaceborne hyperspectral sensors to monitor surface SOC. This study highlights the high accuracy of LSTM with hyperspectral/multispectral data to mitigate a certain level of noise (soil moisture <0.4 m 3 ·m –3 , green leaf area < 0.3 m 2 ·m –2 , plant residue <0.4 m 2 ·m –2 ) for quantifying surface SOC concentration. Forthcoming satellite hyperspectral missions like Surface Biology and Geology (SBG) have a high potential for future global soil carbon monitoring, while high-resolution satellite multispectral fusion data can be an alternative.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2021

Airborne hyperspectral imaging of nitrogen deficiency on crop traits and yield of maize by machine learning and radiative transfer modeling

Nitrogen is an essential nutrient that directly affects plant photosynthesis, crop yield, and biomass production for bioenergy crops, but excessive application of nitrogen fertilizers can cause environmental degradation. To achieve sustainable nitrogen fertilizer management for precision agriculture, there is an urgent need for nondestructive and high spatial resolution monitoring of crop nitrogen and its allocation to photosynthetic proteins as that changes over time. Here, we used visible to shortwave infrared (400–2400 nm) airborne hyperspectral imaging with high spatial (0.5 m) and spectral (3–5 nm) resolutions to accurately estimate critical crop traits, i.e., nitrogen, chlorophyll, and photosynthetic capacity (CO 2 -saturated photosynthesis rate, V max,27 ), at leaf and canopy scales, and to assess nitrogen deficiency on crop yield. We conducted three airborne campaigns over a maize (Zea mays L.) field during the growing season of 2019. Physically based soil-canopy Radiative Transfer Modeling (RTM) and data-driven approaches i.e. Partial-Least Squares Regression (PLSR) were used to retrieve crop traits from hyperspectral reflectance, with ground truth of leaf nitrogen, chlorophyll, V max,27 , Leaf Area Index (LAI), and harvested grain yield. To improve computational efficiency of RTMs, Random Forest (RF) was used to mimic RTM simulations to generate machine learning surrogate models RTM-RF. The results show that prior knowledge of soil background and leaf angle distribution can significantly reduce the ill-posed RTM retrieval. RTM-RF achieved a high accuracy to predict leaf chlorophyll content (R 2 = 0.73) and LAI (R 2 = 0.75). Meanwhile, PLSR exhibited better accuracy to predict leaf chlorophyll content (R 2 = 0.79), nitrogen concentration (R 2 = 0.83), nitrogen content (R 2 = 0.77), and V max,27 (R 2 = 0.69) but required measured traits for model training. We also found that canopy structure signals can enhance the use of spectral data to predict nitrogen related photosynthetic traits, as combining RTM-RF LAI and PLSR leaf traits well predicted canopy-level traits (leaf traits × LAI) including canopy chlorophyll (R 2 = 0.80), nitrogen (R 2 = 0.85) and V max,27 (R 2 = 0.82). Compared to leaf traits, we further found that canopy-level photosynthetic traits, particularly canopy V max,27 , have higher correlation with maize grain yield. This study highlights the potential for synergistic use of process-based and data-driven approaches of hyperspectral imaging to quantify crop traits that facilitate precision agricultural management to secure food and bioenergy production.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2022

A data-centric weak supervised learning for highway traffic incident detection

Using the data from loop detector sensors for near-real-time detection of traffic incidents on highways is crucial to averting major traffic congestion. While recent supervised machine learning methods offer solutions to incident detection by leveraging human-labeled incident data, the false alarm rate is often too high to be used in practice. Specifically, the inconsistency in the human labeling of the incidents significantly affects the performance of supervised learning models. To that end, we focus on a data-centric approach to improve the accuracy and reduce the false alarm rate of traffic incident detection on highways. We develop a weak supervised learning workflow to generate high-quality training labels for the incident data without the ground truth labels, and we use those generated labels in the supervised learning setup for final detection. This approach comprises three stages. First, we introduce a data preprocessing and curation pipeline that processes traffic sensor data to generate high-quality training data through leveraging labeling functions, which can be domain knowledge-related or simple heuristic rules. Second, we evaluate the training data generated by weak supervision using three supervised learning models-random forest, k-nearest neighbors, and a support vector machine ensemble-and long short-term memory classifiers. The results show that the accuracy of all of the models improves significantly after using the training data generated by weak supervision. Third, we develop an online real-time incident detection approach that leverages the model ensemble and the uncertainty quantification while detecting incidents. Finally, we show that our proposed weak supervised learning workflow achieves a high incident detection rate (0.90) and low false alarm rate (0.08).

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2020

Yield strength prediction of high-entropy alloys using machine learning

Yield strength at high temperature is an important parameter in the design and application of high entropy alloys (HEAs). However, the experimental measurement of yield strength at high temperature is quite costly, complicated, and time-consuming. Therefore, it is essential to identify and apply a robust method for the accurate prediction of yield strength at high temperature from the available experimental and simulation data. In this study, for the first time, a machine learning (ML) method based on the regression technique of random forest (RF) regressor is used to predict the yield strength of HEAs at the desired temperature. Further, the yield strengths of MoNbTaTiW and HfMoNbTaTiZr at 800 °C and 1200 °C, are predicted using the RF regressor model. We find that the results are consistent with the experimental reports, showing that the RF regressor model predicts the yield strength of HEAs at the desired temperatures with high accuracy.

36 MATERIALS SCIENCE↗
DOE OSTI2021

Hierarchical Tactile Sensation Integration from Prosthetic Fingertips Enables Multi-Texture Surface Recognition

Multifunctional flexible tactile sensors could be useful to improve the control of prosthetic hands. To that end, highly stretchable liquid metal tactile sensors (LMS) were designed, manufactured via photolithography, and incorporated into the fingertips of a prosthetic hand. Three novel contributions were made with the LMS. First, individual fingertips were used to distinguish between different speeds of sliding contact with different surfaces. Second, differences in surface textures were reliably detected during sliding contact. Third, the capacity for hierarchical tactile sensor integration was demonstrated by using four LMS signals simultaneously to distinguish between ten complex multi-textured surfaces. Four different machine learning algorithms were compared for their successful classification capabilities: K-nearest neighbor (KNN), support vector machine (SVM), random forest (RF), and neural network (NN). The time-frequency features of the LMSs were extracted to train and test the machine learning algorithms. The NN generally performed the best at the speed and texture detection with a single finger and had a 99.2 ± 0.8% accuracy to distinguish between ten different multi-textured surfaces using four LMSs from four fingers simultaneously. The capability for hierarchical multi-finger tactile sensation integration could be useful to provide a higher level of intelligence for artificial hands.

Abd, Moaed A. (ORCID:0000000284954244)↗
DOE OSTI2023

A chemistry-informed hybrid machine learning approach to predict metal adsorption onto mineral surfaces

Historically, surface complexation model (SCM) constants and distribution coefficients (K d ) have been employed to quantify mineral-based retardation effects controlling the fate of metals in subsurface geologic systems. Our recent SCM development workflow, based on the Lawrence Livermore National Laboratory Surface Complexation/Ion Exchange (L-SCIE) database, illustrated a community FAIR data approach to SCM development by predicting uranium(VI)-quartz adsorption for a large number of literature-mined data. Here, we present an alternative hybrid machine learning (ML) approach that shows promise in achieving equivalent high-quality predictions compared to traditional surface complexation models. At its core, the hybrid random forest (RF) ML approach is motivated by the proliferation of incongruent SCMs in the literature that limit their applicability in reactive transport models. Our hybrid ML approach implements PHREEQC-based aqueous speciation calculations; values from these simulations are automatically used as input features for a random forest (RF) algorithm to quantify adsorption and avoid SCM modeling constraints entirely. Named the LLNL Speciation Updated Random Forest (L-SURF) model, this hybrid approach is shown to have applicability to U(VI) sorption cases driven by both ion-exchange and surface complexation, as is shown for quartz and montmorillonite cases. The approach can be applied to reactive transport modeling and may provide an alternative to the costly development of self-consistent SCM reaction databases.

38 RADIATION CHEMISTRY, RADIOCHEMISTRY, AND NUCLEA↗
DOE OSTI2025

Digital twin framework for PIP-II linac: AI-driven multi-scale modeling from ion source to 800 MeV

The PIP-II superconducting linac at Fermilab is designed to deliver multi-megawatt proton beams for neutrino physics and other high-intensity applications. To expedite commissioning and enhance operational reliability, we have developed an EPICS-based data flow framework that seamlessly integrates digital twins (DT) with physical twins (PT). These digital twins comprise high-fidelity beam dynamics models or data-driven surrogate models connected to their physical counterparts through real-time diagnostics and advanced machine-learning algorithms.Central to this framework is Linac_Gen, an accelerated simulation tool that incorporates convolutional neural networks, random forests, and genetic algorithms to provide up to a tenfold speedup in optimizing the accelerator geometry model. An EPICS translator layer ensures interoperability by efficiently mapping lattice parameters across diverse simulation platforms.Our EPICS-based framework supports multiple operational modes—monitoring, passive learning, closed-loop control, and online learning—covering the entire machine lifecycle. By leveraging HPC resources and multi-objective optimization techniques, the digital twin enables adaptive trajectory correction, real-time fault detection, and predictive modeling of beam stability. This comprehensive approach paves the way for robust, high-intensity operation and data-driven accelerator R&D at Fermilab.

Pathak, Abhishek [Fermilab]↗
DOE OSTI2021

Assessing Machine Learning as a Tool to Explain Variance in Deployed Photovoltaic (PV) System Degradation

Degradation remains a large uncertainty in forecasting production for PV plants, creating significant risk for developers and financiers. This study aims to quantify the distribution and drivers of degradation across 10,000 PV systems deployed for distributed or utility generation by training a machine learning model to predict year-over-year degradation rates from metadata characteristics. A combination of K-Means clustering and random forest regressor were found to associate multiple metadata features as potential drivers of degradation, including module characteristics, system design, and climate features. From this, it is inferred that if machine learning is able to find complex patterns between metadata features and system performance loss, such methods can be employed to help developers and financiers make data-informed decisions when estimating long-term energy production forecasts in financial models.

Dunn, Jimmy C.↗
DOE OSTI2026

Virtual Growth of SRF Materials

Niobium's native surface oxide affects SRF cavity and superconducting qubit performance, motivating interest in controlling its crystalline structure. We combine a literature-derived machine-learning analysis with temperature-dependent XRD to study crystalline ordering in Nb2O5. Random Forest models, trained on 74 processing conditions from 17 papers and validated by leave-one-group-out cross-validation, predicted broad crystallinity outcomes well (balanced accuracy 0.809), but struggled with specific polymorph identity (0.577). Annealing temperature was the dominant predictor across all targets; oxygen partial pressure showed negligible importance, reflecting narrow literature coverage rather than physical irrelevance. Temperature-dependent XRD on anodized and H2O2-treated Niobium showed structural evolution consistent with the machine learning predictions. Our model and overall approach provide a data-driven framework for identifying and optimizing conditions that promote crystallization in initially amorphous oxides. This framework can guide the selection of growth and post-annealing conditions for Nb surfaces by narrowing the experimental parameter space, thereby reducing trial-and-error efforts in developing oxide structures relevant to SRF applications.

Tilkin, Anthony [Fermilab]↗
DOE OSTI2026

Virtual Growth of SRF Materials: A Machine Learning Approach to Predict the Crystalline Structural Ordering in Nb Surface Oxides

Niobium's native surface oxide affects SRF cavity and superconducting qubit performance, motivating interest in controlling its crystalline structure. We combine a literature-derived machine-learning analysis with temperature-dependent XRD to study crystalline ordering in Nb2O5. Random Forest models, trained on 74 processing conditions from 17 papers and validated by leave-one-group-out cross-validation, predicted broad crystallinity outcomes well (balanced accuracy 0.809), but struggled with specific polymorph identity (0.577). Annealing temperature was the dominant predictor across all targets; oxygen partial pressure showed negligible importance, reflecting narrow literature coverage rather than physical irrelevance. Temperature-dependent XRD on anodized and H2O2-treated Niobium showed structural evolution consistent with the machine learning predictions. Our model and overall approach provide a data-driven framework for identifying and optimizing conditions that promote crystallization in initially amorphous oxides. This framework can guide the selection of growth and post-annealing conditions for Nb surfaces by narrowing the experimental parameter space, thereby reducing trial-and-error efforts in developing oxide structures relevant to SRF applications.

Tilkin, Anthony [Unlisted, US, IL; Fermilab]↗
DOE OSTI2023

Land surface dynamics and meteorological forcings modulate land surface temperature characteristics

This study examines the effect of land cover, vegetation health, climatic forcings, elevation heat loads, and terrain characteristics (LVCET) on land surface temperature (LST) distribution over West Africa (WA). We employ fourteen machine-learning models, which preserve nonlinear relationships, to downscale LST and other predictands while preserving the geographical variability of WA. Our results showed that the random forest model performs best in downscaling predictands. This is important for the sub-region since it has limited access to mainframes to power multiplex machine-learning algorithms. In contrast to the northern regions, the southern regions consistently exhibit healthy vegetation. Also, areas with unhealthy vegetation coincide with hot LST clusters. The positive Normalized Difference Vegetation Index (NDVI) trends in the Sahel underscore rainfall recovery and subsequent Sahelian greening. The southwesterly winds cause the upwelling of cold waters, lowering LST in southern WA and highlighting the cooling influence of water bodies on LST. Identifying regions with elevated LST is paramount for prioritizing greening initiatives, and our study underscores the importance of considering LVCET factors in urban planning. Topographic slope-facing angles, heat loads, and diurnal anisotropic heat all contribute to variations in LST, emphasizing the need for a holistic approach when designing resilient and sustainable landscapes.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2021

Prediction of DIII-D Pedestal Structure From Externally Controllable Parameters

The sharp increase of pressure at the edge of a high confinement mode (H-mode) plasma, the pedestal, strongly impacts overall plasma performance. Predicting the pedestal is a necessity to control and optimize tokamak operations. Here, an experimental data-driven machine learning (ML) approach is presented that predicts the pedestal heights and widths of electron density (n e ) and electron temperature (T e ) profiles as well as the separatrix ne from externally controllable parameters such as the plasma shape, heating method and power, and gas puff rate and integrated gas puff. The OMFIT framework was used with DIII-D data to efficiently, robustly, and automatically build a database of pedestal parameters to train machine learning models. Database creation was enabled by the search engine tool for DIII-D data, TokSearch, which parallelizes data fetching, enabling fast searches through basic signals of thousands of DIII-D shots and selection of relevant time intervals. Principal Component Analysis (PCA) separated the database into three clusters that represent classes of plasma shapes that are regularly used in DIII-D. The most important parameters for setting the pedestal structure were plasma current (I p ), toroidal magnetic field (B Φ ), neutral beam heating power (P NBI ) and shaping quantities. The Deep Jointly Informed Neural Networks (DJINN) algorithm was applied to identify suitable neural network (NN) architectures that appropriately capture the features of the pedestal database. Separate NNs were implemented for each pedestal parameter, and ensembling methods were used to improve the prediction accuracy and allowed estimation of the prediction uncertainty. The pedestal predictions of the test dataset lie within the measurement uncertainties of the pedestal parameters. The NN outperformed simple Linear Regression (LR) analysis, indicating non-linear dependencies in the pedestal structure. The presented achievements illustrate a promising path for future research, using feature extraction to infer experimental trends and thereby improve pedestal models as well as deploying NN for a fast pedestal prediction in DIII-D scenario development.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2024

Data-Driven Digital Twin for Reliability Assessment of DC/DC Buck Converter

In commercial applications, the operation of DC/DC converters significantly impacts overall system performance and long-term reliability. This study introduces a data-driven digital twin (DT) approach for estimating critical degradation parameters of DC/DC BUCK converter under steady-state condition. Initially, a circuit-level MATLAB/Simulink digital model (DM C ) is refined against a hardware prototype’s switching model dataset using offline particle swarm optimization. The optimized digital model’s steady-state response is then verified with its average model response while varying the duty and load. Subsequently, degradation profiles are imposed on the inductor, capacitor, MOSFET in the DMC. A large dataset is generated from this model, allowing training, validation, and testing of machine learning (ML) models for component health regression tasks. The proposed method employs random forest ML models, achieving impressive regression results with a squared R value as high as 0.99978 and a root mean square error of 4.2× 10 –6 . The method is further validated on a medium power level DC/DC BUCK prototype with varying load conditions, and includes the analysis of MOSFET’s on-resistance under degradation conditions. This data-driven DT method shows promise for identifying parasitic degradation and ohmic loss parameters, enhancing converter reliability assessments in a non-invasive, generalized, and computationally efficient manner.

14 SOLAR ENERGY↗
DOE OSTI2020

Revealing the Formation Energy–Exfoliation Energy–Structure Correlation of MAB Phases Using Machine Learning and DFT

MAB phases are popular as high-temperature materials with high damage tolerance and excellent electrical conductivity that are used to exfoliate two-dimensional (2D) transition metal borides (MBenes); a promising material for developing next- generation nanodevices. In this report, we explore the correlation between the formation energy, exfoliation energy and structural factors of MAB phases using DFT and machine learning. In order to do so, we developed three different machine learning models based on the support vector machine, deep neural network, and random forest regressor to study the stability of the MAB phases by calculating their formation energies. Furthermore, our support vector machine and deep neural network models are capable of predicting the formation energies with mean absolute errors less than 0.1 eV/atom. We demonstrated that the stability of a MAB phase for a given transition metal decreases when the A element changes from Al to Tl. The density functional theory (DFT) re- vealed that M-A and M-B bond strength strongly correlates with the stability of MAB phases. In addition, the exfoliation possibility of 2D MBenes becomes higher when the A element changes from Al to Tl due to weakening of M-A and M-B bonds.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2024

Linac_Gen: Integrating Machine Learning and Particle-in-Cell Methods for Enhanced Beam Dynamics at Fermilab

Here, we introduce Linac_Gen, a tool developed at Fermilab, which combines machine learning algorithms with Particle-in-Cell methods to advance beam dynamics in linacs. Linac_Gen employs techniques such as Random Forest, Genetic Algorithms, Support Vector Machines, and Neural Networks, achieving a tenfold increase in speed for phase-space matching in Linacs over traditional methods, through the use of genetic algorithms. Crucially, Linac_Gen's adept handling of 3D field maps elevates the precision and realism in simulating beam instabilities and resonances, marking a key advancement in the field. Benchmarked against established codes, Linac_Gen demonstrates not only improved efficiency and precision in beam dynamics studies but also in the design and optimization of Linac systems, as evidenced in its application to Fermilab's PIP-II Linac project. This work represents a notable advancement in accelerator physics, marrying ML with PIC methods to set new standards for efficiency and accuracy in accelerator design and research. Linac_Gen exemplifies a novel approach in accelerator technology, offering substantial improvements in both theoretical and practical aspects of beam dynamics.

43 PARTICLE ACCELERATORS↗
DOE OSTI2024

Linac_Gen: integrating machine learning and particle-in-cell methods for enhanced beam dynamics at Fermilab

Here, we introduce Linac_Gen, a tool developed at Fermilab, which combines machine learning algorithms with Particle-in-Cell methods to advance beam dynamics in linacs. Linac_Gen employs techniques such as Random Forest, Genetic Algorithms, Support Vector Machines, and Neural Networks, achieving a tenfold increase in speed for phase-space matching in linacs over traditional methods through the use of genetic algorithms. Crucially, Linac_Gen's adept handling of 3D field maps elevates the precision and realism in simulating beam instabilities and resonances, marking a key advancement in the field. Benchmarked against established codes, Linac_Gen demonstrates not only improved efficiency and precision in beam dynamics studies but also in the design and optimization of linac systems, as evidenced in its application to Fermilab's PIP-II linac project. This work represents a notable advancement in accelerator physics, marrying ML with PIC methods to set new standards for efficiency and accuracy in accelerator design and research. Linac_Gen exemplifies a novel approach in accelerator technology, offering substantial improvements in both theoretical and practical aspects of beam dynamics.

43 PARTICLE ACCELERATORS↗
DOE OSTI2025

A framework to evaluate machine learning crystal stability predictions

The rapid adoption of machine learning in various scientific domains calls for the development of best practices and community agreed-upon benchmarking tasks and metrics. We present Matbench Discovery as an example evaluation framework for machine learning energy models, here applied as pre-filters to first-principles computed data in a high-throughput search for stable inorganic crystals. We address the disconnect between (1) thermodynamic stability and formation energy and (2) retrospective and prospective benchmarking for materials discovery. Alongside this paper, we publish a Python package to aid with future model submissions and a growing online leaderboard with adaptive user-defined weighting of various performance metrics allowing researchers to prioritize the metrics they value most. To answer the question of which machine learning methodology performs best at materials discovery, our initial release includes random forests, graph neural networks, one-shot predictors, iterative Bayesian optimizers and universal interatomic potentials. We highlight a misalignment between commonly used regression metrics and more task-relevant classification metrics for materials discovery. Accurate regressors are susceptible to unexpectedly high false-positive rates if those accurate predictions lie close to the decision boundary at 0 eV per atom above the convex hull. The benchmark results demonstrate that universal interatomic potentials have advanced sufficiently to effectively and cheaply pre-screen thermodynamic stable hypothetical materials in future expansions of high-throughput materials databases.

Riebesell, Janosh↗