Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Active Learning”

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 685 records · Page 38

Using Machine Learning in Adversarial Environments.

Intrusion/anomaly detection systems are among the first lines of cyber defense. Commonly, they either use signatures or machine learning (ML) to identify threats, but fail to account for sophisticated attackers trying to circumvent them. We propose to embed machine learning within a game theoretic framework that performs adversarial modeling, develops methods for optimizing operational response based on ML, and integrates the resulting optimization codebase into the existing ML infrastructure developed by the Hybrid LDRD. Our approach addresses three key shortcomings of ML in adversarial settings: 1) resulting classifiers are typically deterministic and, therefore, easy to reverse engineer; 2) ML approaches only address the prediction problem, but do not prescribe how one should operationalize predictions, nor account for operational costs and constraints; and 3) ML approaches do not model attackers’ response and can be circumvented by sophisticated adversaries. The principal novelty of our approach is to construct an optimization framework that blends ML, operational considerations, and a model predicting attackers reaction, with the goal of computing optimal moving target defense. One important challenge is to construct a realistic model of an adversary that is tractable, yet realistic. We aim to advance the science of attacker modeling by considering game-theoretic methods, and by engaging experimental subjects with red teaming experience in trying to actively circumvent an intrusion detection system, and learning a predictive model of such circumvention activities. In addition, we will generate metrics to test that a particular model of an adversary is consistent with available data.

97 MATHEMATICS AND COMPUTING↗

Exploring Gradient-Based Multi-directional Controls in GANs

Generative Adversarial Networks (GANs) have been widely applied in modeling diverse image distributions. However, despite its impressive applications, the structure of the latent space in GANs largely remains as a black-box, leaving its controllable generation an open problem, especially when spurious correlations between different semantic attributes exist in the image distributions. To address this problem, previous methods typically learn linear directions or individual channels that control semantic attributes in the image space. However, they often suffer from imperfect disentanglement, or are unable to obtain multi-directional controls. Furthermore, in this work, in light of the above challenges, we propose a novel approach that discovers nonlinear controls, which enables multi-directional manipulation as well as effective disentanglement, based on gradient information in the learned GAN latent space. More specifically, we first learn interpolation directions by following the gradients from classification networks trained separately on the attributes, and then navigate the latent space by exclusively controlling channels activated for the target attribute in the learned directions. Empirically, with small training data, our approach is able to gain fine-grained controls over a diverse set of bi-directional and multi-directional attributes, and we showcase its ability to achieve disentanglement significantly better than state-of-the-art methods both qualitatively and quantitatively.

97 MATHEMATICS AND COMPUTING↗

Improving the accessibility and transferability of machine learning algorithms for identification of animals in camera trap images: MLWIC2

Motion-activated wildlife cameras (or “camera traps”) are frequently used to remotely and noninvasively observe animals. The vast number of images collected from camera trap projects has prompted some biologists to employ machine learning algorithms to automatically recognize species in these images, or at least filter-out images that do not contain animals. These approaches are often limited by model transferability, as a model trained to recognize species from one location might not work as well for the same species in different locations. Furthermore, these methods often require advanced computational skills, making them inaccessible to many biologists. We used 3 million camera trap images from 18 studies in 10 states across the United States of America to train two deep neural networks, one that recognizes 58 species, the “species model,” and one that determines if an image is empty or if it contains an animal, the “empty-animal model.” Our species model and empty-animal model had accuracies of 96.8% and 97.3%, respectively. Furthermore, the models performed well on some out-of-sample datasets, as the species model had 91% accuracy on species from Canada (accuracy range 36%–91% across all out-of-sample datasets) and the empty-animal model achieved an accuracy of 91%–94% on out-of-sample datasets from different continents. Our software addresses some of the limitations of using machine learning to classify images from camera traps. By including many species from several locations, our species model is potentially applicable to many camera trap studies in North America. We also found that our empty-animal model can facilitate removal of images without animals globally. We provide the trained models in an R package (MLWIC2: Machine Learning for Wildlife Image Classification in R), which contains Shiny Applications that allow scientists with minimal programming experience to use trained models and train new models in six neural network architectures with varying depths.

59 BASIC BIOLOGICAL SCIENCES↗

Clostridioides difficile Toxin A Remodels Membranes and Mediates DNA Entry Into Cells to Activate Toll-Like Receptor 9 Signaling

Background & Aims: Clostridioides difficile toxin A (TcdA) activates the innate immune response. TcdA co-purifies with DNA. Toll-like receptor 9 (TLR9) recognizes bacterial DNA to initiate inflammation. We investigated whether DNA bound to TcdA activates an inflammatory response in murine models of C difficile infection via activation of TLR9. Methods: We performed studies with human colonocytes and monocytes and macrophages from wild-type and TLR9 knockout mice incubated with TcdA or its antagonist (ODN TTAGGG) or transduced with vectors encoding TLR9 or small-interfering RNAs. Cytokine production was measured with enzyme-linked immunosorbent assay. We studied a transduction domain of TcdA (TcdA 57-80 ), which was predicted by machine learning to have cell-penetrating activity and confirmed by synchrotron small-angle X-ray scattering. Intestines of CD1 mice, C57BL6J mice, and mice that express a form of TLR9 that is not activated by CpG DNA were injected with TcdA, TLR9 antagonist, or both. Enterotoxicity was estimated based on loop weight to length ratios. A TLR9 antagonist was tested in mice infected with C difficile. We incubated human colon explants with an antagonist of TLR9 and measured TcdA-induced production of cytokines. Results: The TcdA 57-80 protein transduction domain had membrane remodeling activity that allowed TcdA to enter endosomes. TcdA-bound DNA entered human colonocytes. TLR9 was required for production of cytokines by cultured cells and in human colon explants incubated with TcdA. TLR9 was required in TcdA-induced mice intestinal secretions and in the survival of mice infected by C difficile. Even in a protease-rich environment, in which only fragments of TcdA exist, the TcdA 57-80 domain organized DNA into a geometrically ordered structure that activated TLR9. Conclusions: TcdA from C difficile can bind and organize bacterial DNA to activate TLR9. TcdA and TcdA fragments remodel membranes, which allows them to access endosomes and present bacterial DNA to and activate TLR9. Rather than inactivating the ability of DNA to bind TLR9, TcdA appears to chaperone and organize DNA into an inflammatory, spatially periodic structure.

59 BASIC BIOLOGICAL SCIENCES↗

Machine learning model to project the impact of COVID-19 on U.S. motor gasoline demand

Owing to the global lockdowns that resulted from the COVID-19 pandemic, fuel demand plummeted and the price of oil futures went negative in April 2020. Robust fuel demand projections are crucial to economic and energy planning and policy discussions. Here we incorporate pandemic projections and people's resulting travel and trip activities and fuel usage in a machine-learning-based model to project the US medium-term gasoline demand and study the impact of government intervention. We found that under the reference infection scenario, the US gasoline demand grows slowly after a quick rebound in May, and is unlikely to fully recover prior to October 2020. Under the reference and pessimistic scenario, continual lockdown (no reopening) could worsen the motor gasoline demand temporarily, but it helps the demand recover to a normal level quicker. Under the optimistic infection scenario, gasoline demand will recover close to the non-pandemic level by October 2020. The COVID 19 pandemic and consequent lockdown has had a substantial impact on mobility and therefore fuel demand and it is not clear when demand will recover. Ou et al. use a machine learning model that integrates health recovery scenarios to project the near-term future of gasoline demand.

COVID-19 pandemic↗

Adaptive learning-driven high-throughput synthesis of oxygen reduction reaction Fe–N–C electrocatalysts

Reducing human reliance on inefficient energy systems and fossil fuels has become more urgent due to the consequences of global climate change. However, traditional trial-and-error approaches have hampered our ability to accelerate the discovery and implementation of functional materials for efficient energy conversion devices, such as polymer electrolyte fuel cells (PEFCs). To address this, we develop an adaptive learning framework that integrates machine learning and state-of-the-art capabilities in high-throughput synthesis to achieve expedited optimization of iron-nitrogen-carbon PEFC oxygen reduction reaction (ORR) electrocatalysts. We use statistical inference, uncertainty quantification, and global optimization to build a computational design-of-experiment tool that identifies the optimum compositions to be investigated next to reduce the demands placed on experimental materials discovery. We benchmark the ability of the proposed strategy to discover optimum catalyst synthesis conditions in a six-dimensional search space when starting with a thirty-six-sample database. By following the adaptive learning strategy, we synthesize fourteen new catalysts from approximately ten billion unique compositions and discover four catalysts that outperform all original samples. The best machine learning-optimized catalyst is 33% more active than the highest-performing one in the initial database, showing an ORR activity seven times larger than those typically reported for the same class of materials.

36 MATERIALS SCIENCE↗

Coupling High-Throughput Experiments and Regression Algorithms to Optimize PGM-Free ORR Electrocatalyst Synthesis

Over the past decades, significant improvement has been achieved in the performance of platinum group metal-free (PGM-free) materials as an alternative to Pt-based electrocatalysts for oxygen reduction reaction (ORR). However, further progress in ORR activity requires evaluation of precursors and synthesis approaches. In response to this challenge, we generated a first of its kind experimental data set of 36 samples using high-throughput synthesis and activity measurements. Several control parameters (e.g., Fe precursor identity, the precursor content, and pyrolysis temperature) were varied. We then developed several state-of-the-art machine learning (ML) based regression models to predict ORR activity, dependent on selected synthesis variables. Through an iterative algorithm, higher prediction accuracy (smaller root-mean-square error) was achieved. We identified that gradient boosting regression (GBR) and support vector regression (SVR), among several methods, work best for this data set. Aided by our ML-based surrogate models, we decided to alter catalyst synthesis conditions, which resulted in a 36% increase in measured ORR activity in comparison to the maximum ORR mass activity value of 21.9 A/g catalyst in the original data set. Overall, this combined experiment and machine learning approach represents a promising path forward toward developing highly efficient next-generation ORR electrocatalysts and, more generally, functional materials.

25 ENERGY STORAGE↗

SU-G-BRA-10: Marker Free Lung Tumor Motion Tracking by An Active Contour Model On Cone Beam CT Projections for Stereotactic Body Radiation Therapy of Lung Cancer

Purpose: To develop a novel strategy to extract the lung tumor motion from cone beam CT (CBCT) projections by an active contour model with interpolated respiration learned from diaphragm motion. Methods: Tumor tracking on CBCT projections was accomplished with the templates derived from planning CT (pCT). There are three major steps in the proposed algorithm: 1) The pCT was modified to form two CT sets: a tumor removed pCT and a tumor only pCT, the respective digitally reconstructed radiographs DRRtr and DRRto following the same geometry of the CBCT projections were generated correspondingly. 2) The DRRtr was rigidly registered with the CBCT projections on the frame-by-frame basis. Difference images between CBCT projections and the registered DRRtr were generated where the tumor visibility was appreciably enhanced. 3) An active contour method was applied to track the tumor motion on the tumor enhanced projections with DRRto as templates to initialize the tumor tracking while the respiratory motion was compensated for by interpolating the diaphragm motion estimated by our novel constrained linear regression approach. CBCT and pCT from five patients undergoing stereotactic body radiotherapy were included in addition to scans from a Quasar phantom programmed with known motion. Manual tumor tracking was performed on CBCT projections and was compared to the automatic tracking to evaluate the algorithm accuracy. Results: The phantom study showed that the error between the automatic tracking and the ground truth was within 0.2mm. For the patients the discrepancy between the calculation and the manual tracking was between 1.4 and 2.2 mm depending on the location and shape of the lung tumor. Similar patterns were observed in the frequency domain. Conclusion: The new algorithm demonstrated the feasibility to track the lung tumor from noisy CBCT projections, providing a potential solution to better motion management for lung radiation therapy.

60 APPLIED LIFE SCIENCES↗

Generalized Brønsted‐Evans‐Polanyi Relationships for Reactions on Metal Surfaces from Machine Learning

Abstract Brønsted‐Evans‐Polanyi (BEP) relationships, i. e., a linear scaling between reaction and activation energies, lie at the core of computational design of heterogeneous catalysts. However, BEPs are not general and often require reparameterization for each class of reactions. Here we construct generalized BEPs (gBEPs), which can predict activation energies for a diverse dataset of reactions of C, O, N and H containing molecules on metal surfaces. In a first step we develop a set of descriptors based on scaling relationships that can capture the change in chemical identity of reactants during the reaction. Subsequently, we use the reaction energy, these descriptors and a single descriptor for the surface structure to parameterize machine learning based regression approaches for the prediction of activation energies. The best approach we developed shows a Mean Absolute Error (MAE) of 0.11 eV for the training set (80 % of the data set) and 0.23 eV for the test set (20 % of the data set). The methodology presented here allows to calculate activation energies within fractions of seconds on a typical personal computer and due to its generality, accuracy and simplicity in application it might prove to be useful in transition metal catalyst design.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

WE-E-BRF-01: The ESTRO-AAPM Joint Symposium On Imaging for Proton Treatment Planning and Guidance

In this first inaugural joint ESTRO-AAPM session we will attempt to provide some answers to the problems encountered in the clinical application of particle therapy. Indeed the main advantage is that the physical properties of ion beams offer high ballistic accuracy for tightly conformal irradiation of the tumour volume, with excellent sparing of surrounding healthy tissue and critical organs, This also its Achilles' heel calling for an increasing role of imaging to ensure safe application of the intended dose to the targeted area during the entire course of fractionated therapy. We have three distinguished speakers addressing possible solutions. Katia Parodi (Ludwig Maximilians University, Munich, Germany) To date, Positron Emission Tomography (PET) is the only technique which has been already clinically investigated for in-vivo visualization of the beam range during or shortly after ion beam delivery. The method exploits the transient amount of β{sup 2}-activity induced in nuclear interactions between the primary beam and the irradiated tissue, depending on the ion beam species, the tissue elemental composition and physiological properties (in terms of biological clearance), as well as the time course of irradiation and imaging. This contribution will review initial results, ongoing methodological developments and remaining challenges related to the clinical usage of viable but often suboptimal instrumentation and workflows of PET-based treatment verification. Moreover, it will present and discuss promising new detector developments towards next-generation dedicated PET scanners relying on full-ring or dual-head designs for in-beam quasi real-time imaging. Denis Dauvergne (Institut de Physique Nucleaire de Lyon, Lyon, France) Prompt gamma radiation monitoring of hadron therapy presents the advantage of real time capability to measure the ion range. Both simulations and experiments show that millimetric verification of the range can be achieved at the pencil beam scale for active proton beam delivery in homogenous targets. The development of gamma cameras, that has been studied by several groups worldwide over the last years, now reaches - for some of them - the stage of being applicable in clinical conditions, with real size prototypes and count rate capability matching the therapeutic beam intensities. We will review the different concepts of gamma cameras, the advantages and limitations of this method, and the main challenges that should still be overcome before the widespread of prompt gamma quality assurance for proton and hadrontherapy. Jon Kruse (Mayo Clinic, Rochester, MN, USA) Treatment simulation images for proton therapy are used to determine proton stopping power and range in the patient. This talk will discuss the careful control of CT numbers and conversion of CT number to stopping power required in proton therapy. Imaging for treatment guidance of proton therapy also presents unique challenges which will be addressed. Among them are the enhanced relationship between internal anatomy changes and dosimetry, the need for imaging to support adaptive planning protocols, and high operational efficiency. Learning Objectives: To learn about the possibilities of using activation products to determine the range of particle beams in a patient treatment setting To be informed on an alternative methodology using prompt gamma detectors To understand the impact of the accuracy of the knowledge of the patient information with respect to the delivered treatment.

60 APPLIED LIFE SCIENCES↗

DOE FAIR Surrogate Benchmarks Supporting AI and Simulation Research (SBI Surrogate Benchmark Initiative) (Final Report)

Computational Science is being revolutionized by integrating AI and simulation and, in particular, by deep learning surrogate models that can replace all or part of traditional large‐scale HPC computations. Such surrogates can achieve remarkable performance improvements, as much as several orders of magnitude, and save both compute time and energy. The Surrogate Benchmark Initiative (SBI) project creates a community repository and FAIR (Findable, Accessible, Interoperable, and Reusable) data ecosystem for HPC application surrogate benchmarks. The SBI team comes from Argonne National Laboratory (ANL), Indiana University (IU), Rutgers University, the University of Tennessee, Knoxville (UTK), and the University of Virginia(UVA). SBI repositories include data, code, and all relevant collateral artifacts, that the science and engineering community needs to use and reuse these data sets and surrogates. SBI repositories generate active research from both participants in SBI and the broader AI and domain science communities. This project develops surrogates that use several different neural nets to learn and quickly infer the results of simulations and data systems and capture them as surrogate benchmarks with a rich set of metadata, covering. Data; Model; Metrics specification; Machine specification; Science, Speed, Power Results, We research FAIR metadata for these benchmarks. We develop application surrogate examples as benchmarks across many fields (ANL, UTK, IU, UVA). We also study non Surrogate benchmarks that have many common features and similar issues regarding FAIRness. We work with MLCommons (UVA, UTK), which is a major machine learning benchmarking activity where we get metadata ontologies, software, and benchmarks, benchmarks have datasets, models, and metadata, and they need a technical framework developed by UTK and Rutgers and deployed by UVA. We study features of Surrogates, including performance, training set size, and uncertainty quantification (Rutgers, UVA and IU).

97 MATHEMATICS AND COMPUTING↗

Final Demolition and Disposition of 209-E Critical Mass Laboratory - 12267

The 209-E Critical Mass Laboratory was constructed in 1960 to provide a heavy shielded reactor room where quantities of plutonium or uranium in solution could be brought to near-critical configurations under carefully controlled and monitored conditions. In the late 1980's, the responsible contractor, Pacific Northwest National Laboratory (PNNL), was directed by the Department of Energy (DOE) to prepare the facility for unoccupied status. The facility was demolished under a Removal Action Work Plan pursuant to the Comprehensive Environmental Response, Compensation, and Liability Act of 1980 (CERCLA). The funding for this project was provided by the American Recovery and Reinvestment Act (ARRA). The primary rooms of concern with regards to contamination in 209-E facility, which is over 9,000 square feet, are the criticality assembly room (CAR), the mix room, and the change room. The CAR contained two reactor hoods (HO-140 and HO-170), which each had a high efficiency particulate air (HEPA) filter system. The CAR contained 13 tanks ranging from 38 L (10 gal) to 401 L (106 gal). Tanks TK-109 and TK-110 are below grade, and were removed as part of this demolition and disposition remedy. Nonradiological and radiological hazardous substances were removed, decontaminated, or fixed in place, prior to demolition. Except for the removal of below grade tanks TK-109 and TK-110, the facility was demolished to slab-on-grade. PNNL performed stabilization and deactivation activities that included removal of bulk fissile material and chemicals, flushing tanks, stabilizing contamination within gloveboxes and hoods, and packaging and removing waste. The removal of the contaminated plutonium equipment and materials from the 209E facility presented a number of challenges similar in nature to those associated with the inventory reduction and cleanup activities at the Plutonium Finishing Plant. Although there were no bulk fissile materials or chemicals within the facility, there were residual radiological materials (isotopes of plutonium and americium) in the tanks and hoods. The complexity of the remedy was present because of the various configurations of the tanks and hoods, combined with the residual contaminants. Because of the weight and dimensional configuration, size reduction of the slab tanks, as well as removal and disposal of the different material used for moderation and absorption, were two examples of challenges that were resolved to complete the remedy. One of the key methods developed and implemented at the facility was the design and construction of a shroud to allow the cutting of the Pu contaminated tanks. The shroud design, development and implementation at the 209E Project was an example of enhanced work planning and task hazards analysis with worker involvement. This paper will present the lessons learned from the 209E facility inventory reduction activities including the shroud and other methodologies used. The initial Lessons Learned discussion for this project was scheduled for late January 2012. This facility is the first open-air demolition of a highly contaminated plutonium-contaminated facility accomplished by CH2M Hill under the Plateau Remediation Contract. The demolition was completed without spread of contamination to the workers and the surrounding area. As with any project of this complexity, there are significant accomplishments, as well as experience that can be applied to future demolition of plutonium-contaminated facilities on the Hanford Site. These experiences will be documented at a later date. (authors)

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗

Subtask 1.5 – CO2 Injection Monitoring with an Optimized Scalable, Automated, Semipermanent Seismic Array

The scalable, automated, semipermanent seismic array (SASSA) method is a flexible and relatively cost-effective surface geophysical method for regular time-lapse monitoring of the movement of injected carbon dioxide (CO2) in a reservoir for CO2 enhanced oil recovery (EOR) or geologic CO2 storage operations. It has the advantages of a low-environmental-footprint while monitoring regions of a reservoir from the surface without the need for a regular grid distribution of receivers. Automated data collection is possible. As only time-lapse amplitude changes at the reservoir level due to CO2 movement within the reservoir are monitored, the turnaround time to deliver results from the SASSA method can be short, without the need for long, time-consuming data-processing workflows. As data is collected and processed, incremental information can be provided to the field operator. The Energy & Environmental Research Center (EERC) conducted a SASSA field test from September 2018 to November 2020 in a portion of the Bell Creek Field in Montana, which implemented new CO2 EOR field activities during the study period. Lessons learned from a proof-of-concept study were incorporated to improve the data quality of the SASSA method and demonstrate the viability of the technology. The EERC implemented several enhancements to improve data quality, including 1) an iterative survey design, which allowed placing the receivers in strategic locations where the movement of the CO2 in the reservoir could be tracked with minimum interference by the cultural noise in the study area; 2) the use of powerful seismic sources in the form of surface orbital vibrators, and 3) data acquisition during optimal periods. History-matched reservoir simulation was performed to predict gas saturation and pressure response induced by CO2 injection in the study area. The results were compared with the SASSA-measured responses to CO2 injection as a partial validation technique. A match between the two methods was observed for most of the SASSA points predicted to have intersected a CO2 saturation change. The validated results provide confidence that the SASSA method can be used independently as a CO2 saturation monitoring technique. As data are collected and processed, incremental information can be provided to the field operator. The critical components of the SASSA workflow for a successful application of the method are the following: Iterative survey design with information about CO2 injection activities from the oilfield operator. A detailed CO2 injection plan is the key driver to select the strategic monitoring location of the SASSA sensors. After this information is incorporated in the initial distribution of sources and receivers in the study area, high-resolution satellite images are used to identify ground locations not affected by cultural noise sources, such as power lines, pipelines/flow lines, or roadways. In the next iteration of the survey design, a scouting trip to the study area is needed to understand more details of the noise sources identified in the previous step and the intensity of the field activities that can also generate noise during the monitoring. Integrating the information from the scouting trip into the survey design to select the optimum source and receiver locations is the final step. Noise attenuation. The variety of noise types during seismic monitoring of an oil field is enormous. Tailored noise characterization and processing at a node-by-node level can enhance the performance and sensitivity of the SASSA technique. Future advancements that could improve the efficiency and application of the SASSA technology include: Gaining a better understanding of the noise field produced by the seismic source to aid the choice of receiver location. Surface noise from the source can overwhelm the small signal changes due to CO2 that the SASSA method measures. Improved data-processing workflow to automatically analyze and adapt to dynamic noise conditions associated with industrial settings. This subtask was funded through the EERC–DOE Joint Program on Research and Development for Fossil Energy-Related Resources Cooperative Agreement No. DE- FE0024233.

29 ENERGY PLANNING, POLICY, AND ECONOMY↗

In silico modeling to predict drug-induced phospholipidosis

Drug-induced phospholipidosis (DIPL) is a preclinical finding during pharmaceutical drug development that has implications on the course of drug development and regulatory safety review. A principal characteristic of drugs inducing DIPL is known to be a cationic amphiphilic structure. This provides evidence for a structure-based explanation and opportunity to analyze properties and structures of drugs with the histopathologic findings for DIPL. In previous work from the FDA, in silico quantitative structure–activity relationship (QSAR) modeling using machine learning approaches has shown promise with a large dataset of drugs but included unconfirmed data as well. In this study, we report the construction and validation of a battery of complementary in silico QSAR models using the FDA's updated database on phospholipidosis, new algorithms and predictive technologies, and in particular, we address high performance with a high-confidence dataset. The results of our modeling for DIPL include rigorous external validation tests showing 80–81% concordance. Furthermore, the predictive performance characteristics include models with high sensitivity and specificity, in most cases above ≥ 80% leading to desired high negative and positive predictivity. These models are intended to be utilized for regulatory toxicology applied science needs in screening new drugs for DIPL. - Highlights: • New in silico models for predicting drug-induced phospholipidosis (DIPL) are described. • The training set data in the models is derived from the FDA's phospholipidosis database. • We find excellent predictivity values of the models based on external validation. • The models can support drug screening and regulatory decision-making on DIPL.

60 APPLIED LIFE SCIENCES↗

Machine Learning with Gradient-Based Optimization of Nuclear Waste Vitrification with Uncertainties and Constraints

Gekko is an optimization suite in Python that solves optimization problems involving mixed-integer, nonlinear, and differential equations. The purpose of this study is to integrate common Machine Learning (ML) algorithms such as Gaussian Process Regression (GPR), support vector regression (SVR), and artificial neural network (ANN) models into Gekko to solve data based optimization problems. Uncertainty quantification (UQ) is used alongside ML for better decision making. These methods include ensemble methods, model-specific methods, conformal predictions, and the delta method. An optimization problem involving nuclear waste vitrification is presented to demonstrate the benefit of ML in this field. ML models are compared against the current partial quadratic mixture (PQM) model in an optimization problem in Gekko. GPR with conformal uncertainty was chosen as the best substitute model as it had a lower mean squared error of 0.0025 compared to 0.018 and more confidently predicted a higher waste loading of 37.5 wt% compared to 34 wt%. The example problem shows that these tools can be used in similar industry settings where easier use and better performance is needed over classical approaches. Future works with these tools include expanding them with other regression models and UQ methods, and exploration into other optimization problems or dynamic control.

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗

FAIR Surrogate Benchmarks Supporting AI and Simulation Research (Final Report)

Computational Science is being revolutionized by integrating AI and simulation and, in particular, by deep learning surrogate models that can replace all or part of traditional large‐scale HPC computations. Such surrogates can achieve remarkable performance improvements, as much as several orders of magnitude, and save both compute time and energy. The Surrogate Benchmark Initiative (SBI) project creates a community repository and FAIR (Findable, Accessible, Interoperable, and Reusable) data ecosystem for HPC application surrogate benchmarks. The SBI team comes from Argonne National Laboratory (ANL), Indiana University (IU), Rutgers University, the University of Tennessee, Knoxville (UTK), and the University of Virginia (UVA). SBI repositories include data, code, and all relevant collateral artifacts that the science and engineering community need to use and reuse these data sets and surrogates. SBI repositories generate active research from both the participants in SBI and the broad community of AI and domain scientists. This project develops surrogates that use several different neural nets to learn and quickly infer the results of simulations and data systems and captures them as surrogate benchmarks with a rich set of metadata covering: Data; Model; Metrics specification; Machine specification; and Science, Speed, and Power Results. We research FAIR metadata for these benchmarks. We develop application surrogate examples as benchmarks across many fields (ANL, UTK, IU, UVA). We also study non-Surrogate benchmarks that have many common features and similar issues as regards FAIRness. We work with MLCommons (UVA, UTK), which is a major machine learning benchmarking activity where we get metadata ontologies, software, and benchmarks, Benchmarks have datasets, models, and metadata and they need a technical framework developed by UTK and Rutgers and deployed by UVA. We study features of Surrogates including performance, training set size, and uncertainty quantification (Rutgers, UVA and IU).

97 MATHEMATICS AND COMPUTING↗

Post-Disaster Microgrid Formation for Enhanced Distribution System Resilience

This paper proposes a deep reinforcement learning (DRL) based approach for post-disaster critical load restoration in active distribution systems to form microgrids through network reconfiguration to minimize critical load curtailments. Distribution networks are represented as graph networks, and optimal network configurations with microgrids are obtained by searching for the optimal spanning forest. The constraints to the research question being explored are the radial topology and power balance. Unlike existing analytical and population-based approaches, which necessitate the repetition of entire analyses and computation for each outage scenario to find the optimal spanning forest, the proposed approach, once properly trained, can quickly determine the optimal, or near-optimal, spanning forest even when outage scenarios change. When multiple lines fail in the system, the proposed approach forms microgrids with distributed energy resources in active distribution systems to reduce critical load curtailment. The proposed DRL-based model learns the action-value function using the REINFORCE algorithm, which is a model-free reinforcement learning technique based on stochastic policy gradients. A case study was conducted on a 33-node distribution test system, demonstrating the effectiveness of the proposed approach for post-disaster critical load restoration.

active distribution systems↗

Effect of Solvent on the Local Structure, Dynamics, and Vibrational Density of States in Sn-BEA Zeolite

Lewis acid zeolites are attractive catalysts for epoxidation and biomass valorization, as they are highly active and selective in the liquid phase and can operate at or near ambient conditions. While a rich experimental literature exists on liquid-phase Lewis acid zeolite catalysis, our understanding of the molecular organization and solvent dynamics in the vicinity of Lewis acid sites with differing metal site speciation remains limited. In this work, we investigate the molecular coordination and diffusion of two common solvents (methanol and water) around the closed and open Sn-BEA zeolite active sites using molecular dynamics simulations with a machine-learned interatomic potential trained on ab initio molecular dynamics trajectories. Molecular dynamics simulations reveal that introducing active sites significantly enhances local order in the first and second solvation shells compared to the pure silica case. For methanol, both closed and open active sites are singly coordinated, while more than two water molecules coordinate the open site. In contrast to methanol, we observed that water molecules dissociate, leading to the formation of additional Sn-OH and silanol groups away from the active site. The diffusion coefficients of water and methanol are functions of the solvent population in the pore. Here, our work provides insights into how active site speciation in Lewis acid zeolites affects solvent coordination, diffusion, and vibrational signature. This information is foundational for catalyst design and optimization of liquid-phase catalytic processes in zeolites. It also demonstrates the suitability of machine-learned interatomic potentials for modeling reactive systems, enabling sufficiently long trajectories for appropriate statistical averaging.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗