Search NASA⌕ Search

SEARCH · Search NASA

Results for “learning classifiers”

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.

520 records · Page 29
All sourcesNASA NTRSDOE OSTI
DOE OSTI2024

Data Science Shows that Entropy Correlates with Accelerated Zeolite Crystallization in Monte Carlo Simulations

We have performed a data science study of Monte Carlo simulation trajectories to understand factors that can accelerate formation of zeolite nanoporous crystals, a process that can take days or even weeks. In previous work, Monte Carlo simulations predicted and experiments confirmed that using a secondary organic structure-directing agent (OSDA) accelerates crystallization of all-silica LTA zeolite, with experiments finding a three-fold speedup [PCCP 24, 142-148 (2022)]. However, it remains unclear what physical factors cause the speed-up. Here, we apply data science to analyze the simulation trajectories to discover what drives accelerated zeolite crystallization in Monte Carlo going from a one-OSDA synthesis (1OSDA) to a two-OSDA version (2OSDA). We encoded simulation snapshots using the Smooth Overlap of Atomic Positions approach, which represents all 2- and 3-body correlations within a given cutoff distance. Principal component analyses failed to discriminate datasets of structures from 1OSDA and 2OSDA simulations, while the Support Vector Machine (SVM) approach succeeded at classifying such structures with an area-under-curve (AUC) score of 0.99 (where AUC = 1 is a perfect classification) with all 3-body correlations, and as high as 0.94 with only 2-body correlations. SVM decision functions reveal relatively broad / narrow histograms for 1OSDA / 2OSDA datasets, suggesting that the two simulations differ strongly in information heterogeneity. Informed by these results, we performed pair (2-body) entropy calculations during crystallization, resulting in entropy differences that semi-quantitatively account for the speedup observed in the previous Monte Carlo simulations. We conclude that altering synthesis conditions in ways that substantially changes the entropy of labile silica networks may accelerate zeolite crystallization, and we discuss possible approaches for achieving such acceleration.

77 NANOSCIENCE AND NANOTECHNOLOGY↗
DOE OSTI2026

Revealing the Hidden Third Dimension of Point Defects in Two-Dimensional MXenes

Point defects govern many important functional properties of two-dimensional (2D) materials. However, resolving the three-dimensional (3D) arrangement of these defects in multi-layer 2D materials remains a fundamental challenge, hindering rational defect engineering. Here, we overcome this limitation using an artificial intelligence-guided electron microscopy workflow to map the 3D topology and clustering of atomic vacancies in Ti3C2TX MXene. Our approach reconstructs the 3D coordinates of vacancies across hundreds of thousands of lattice sites, generating robust statistical insight into their distribution that can be correlated with specific synthesis pathways. This large-scale data enables us to classify a hierarchy of defect structures-from isolated vacancies to nanopores-revealing their preferred formation and interaction mechanisms, as corroborated by molecular dynamics simulations. This work provides a generalizable framework for understanding and ultimately controlling point defects across large volumes, paving the way for the rational design of defect-engineered functional 2D materials.

2D materials↗
NASA NTRS2025

A Preliminary Study on the Feasibility of Large Language Models for Detecting Micro-Behaviors Among Team Members in Space Missions

Large-language models (LLMs) have been recently used for spoken language understanding (SLU) to infer meaning and semantics from speech in tasks such as speaker intent and sentiment classification. Due to being trained on large amounts of data, and their ability to understand context and relationships between words, LLMs are competent, enabling them to generalize across tasks without requiring many task-specific training samples. This research examines the feasibility of few-shot learning in LLMs for detecting subtle, brief, and possibly unconscious interactions between team members, called ``micro-behaviors," and provides insights into the appropriate design of LLMs for this task. Our data came from 5 teams participating in a 45-day mission at the US National Aeronautics and Space Administration’s (NASA) Human Exploration Research Analog (HERA). More specifically we used data collected from team interaction battery (TIB) tasks teams performed five times in-mission which comprise an average 1.5 hours of conversation data per day. Micro-behaviors were coded according to an adapted version of Smith & Griffins (2022) theoretical framework in terms of Violation (i.e., presence of valenced behavior, uplifting/positive or discouraging/negative), Intensity (i.e., force of behavior in terms of how uplifting or discouraging is the behavior), and Intent (i.e., motive of the behavior in terms of whether it was deliberate or unintentional). We explore the ability of LLMs to detect the presence and intensity of micro-behaviors. We examine employing and fine-tuning readily available LLMs (i.e., RoBERTa, DistilBERT), as well as prompting state-of-the-art sequence classification models (i.e., Llama-2, Llama-3). In a total of 13,058 conversational turns (17.8% uplifting, 3.3% discouraging, 75.76% neutral, 3.14% nulls), we compute the macro F1-score of the 3-way micro-behavior classification task (i.e., classifying among uplifting, discouraging, and neutral; 33% chance). Results indicate that the RoBERTa model achieves a F1-score of 36.2% (uplift: 43.3% precision (P), 15.1% recall (R); discourage: 20% P, 0.5% R). These results significantly improve when we augment the data via paraphrasing in the RoBERTa model, reaching a 41.2% macro F1-score (uplift: 37.7% P, 86.3% R; discourage: 3.5% P, 1.8% R). Finally, the Llama-2 model with 3-shot prompting yields 38% macro F1-score (uplift: 28.7% P, 20% R; discourage: 7.2% P, 18% R), which is slightly better compared to the RoBERTa model without data augmentation, highlighting the effectiveness of sequence classification models in detecting minority classes with a small sample size. Findings indicate that LLMs hold potential to detect subtle behaviors in conversations, which could be valuable in assessing team behavior in space exploration missions. Future studies will evaluate the performance of different LLM prompting strategies or fine-tuning methods.

Ankush Raut↗
NASA NTRSDate not supplied

Mapping Inundation from Hurricane Florence (2018) with L-Band Synthetic Aperture Radar, Commercial Imagery, and Ancillary Data via Machine Learning Classification

During and after flooding events, mapping the extent of floodwaters aids in the distribution of resources, recovery efforts, and damage assessment practices. Development of a land cover classification system focused on mapping inundation after major hurricane events using synthetic aperture radar (SAR) data could allow for the production of near-real-time inundation mapping, enabling government and emergency response entities to get a preliminary idea of a developing situation. Complimentary optical and SAR images from domestic and foreign entities are brought together through activations of the International Charter: Space and Major Disasters to support response efforts, from true-color, near-infrared, and thermal remote sensing data obtained by NASA, NOAA, and international satellites to the collection of high-resolution true color aerial photography by NOAA and the National Geodetic Survey. In response to Hurricane Florence of 2018, NASA JPL collected numerous swaths of quad-pol L-band SAR data with the Uninhabited Aerial Vehicle Synthetic Aperture Radar (UAVSAR) instrument observing the record-setting river stages across North and South Carolina. The resulting fully-polarized SAR images allow for mapping of inundation extent at a high spatial resolution with a unique advantage over optical imaging stemming from the sensor’s ability to penetrate cloud cover and dense vegetation. In this study, true-color NOAA aerial and commercial satellite imagery are used in conjunction with four UAVSAR data swaths centered on the Lumberton and Cape Fear River basins in southeastern North Carolina to develop a Random Forest classification model focused on mapping open water and floodwater otherwise obscured by vegetation or lingering cloud cover. Ancillary building footprint, transportation route, and population data will also be incorporated into the classification scheme to estimate the societal impacts of flooding based on the proximity of features to detected inundation. Preliminary results from the Hurricane Florence case study will be discussed in addition to the limitations of available validation data for assessment of the classifier’s accuracy.

Alexander M Melancon↗
DOE OSTI2024

GrainGNN: A dynamic graph neural network for predicting 3D grain microstructure

We propose GrainGNN, a surrogate model for the evolution of polycrystalline grain structure under rapid solidification conditions in metal additive manufacturing. High fidelity simulations of solidification microstructures are typically performed using multicomponent partial differential equations (PDEs) with moving interfaces. The inherent randomness of the PDE initial conditions (grain seeds) necessitates ensemble simulations to predict microstructure statistics, e.g., grain size, aspect ratio, and crystallographic orientation. Here, currently such ensemble simulations are prohibitively expensive and surrogates are necessary.In GrainGNN, we use a dynamic graph to represent interface motion and topological changes due to grain coarsening. We use a reduced representation of the microstructure using hand-crafted features; we combine pattern finding and altering graph algorithms with two neural networks, a classifier (for topological changes) and a regressor (for interface motion). Both networks have an encoder-decoder architecture; the encoder has a multi-layer transformer long-short-term-memory architecture; the decoder is a single layer perceptron.We evaluate GrainGNN by comparing it to high-fidelity phase field simulations for in-distribution and out-of-distribution grain configurations for solidification under laser power bed fusion conditions. GrainGNN results in 80%–90% pointwise accuracy; and nearly identical distributions of scalar quantities of interest (QoI) between phase field and GrainGNN simulations compared using Kolmogorov-Smirnov test. GrainGNN's inference speedup (PyTorch on single x86 CPU) over a high-fidelity phase field simulation (CUDA on a single NVIDIA A100 GPU) is 150×–2000× for 100-initial grain problem. Further, using GrainGNN, we model the formation of 11,600 grains in 220 seconds on a single CPU core.

36 MATERIALS SCIENCE↗
DOE OSTI2026

Unsupervised anomaly detection in MeV ultrafast electron diffraction

MeV ultrafast electron diffraction (MUED) is a pump-probe technique used to study the dynamic structural evolution of materials. An ultrashort laser pulse triggers structural changes, which are then probed by an ultrashort relativistic electron beam. To overcome low signal-to-noise ratios, diffraction patterns are averaged over thousands of shots. However, shot-to-shot instabilities in the electron beam can distort individual patterns, introducing uncertainty. Improving MUED accuracy requires detecting and removing these anomalous patterns from large datasets. In this work, we developed a fully unsupervised methodology for the detection of anomalous diffraction patterns. Using a convolutional autoencoder, we calculate the reconstruction mean squared error of the diffraction patterns. Based on the statistical analysis of this error, we provide the user an estimation of the probability that the pattern is normal, which also allows a posterior visual inspection of the images that are difficult to classify. This method has been trained with only 100 diffraction patterns and tested on 1521 patterns, resulting in a false positive rate between 0.2% and 0.4%, with a training time of 10 s per image and a test time of about 1 s per image. Here, the proposed methodology can also be applied to other diffraction techniques in which large datasets are collected that include faulty images due to instrumental instabilities.

43 PARTICLE ACCELERATORS↗
DOE OSTI2025

Data-Efficient Methods for Determining Flory–Huggins χ Parameters in Multicomponent Polymer Formulations

Polymer formulations are essential in diverse applications including personal care products, coatings, paints, adhesives, and plastic materials. Designing these formulations requires navigating large, complex design spaces, where phase and self-assembly behavior critically impact performance. The Flory–Huggins χ parameter, which quantifies segmental miscibility, is widely used to parametrize the excess free energy of mixing in formulation models. In this work, we introduce two data-efficient, top-down methods for estimating χ parameters using the Random Phase Approximation (RPA): (i) Boundary Nonlinear Regression (Boundary-NLR), which fits theoretical spinodal boundaries to experimental phase boundaries, and (ii) Surrogate Model Inverse Parameter Estimation (SMIPE), which uses a Gaussian Process Classifier to fit sparse phase maps via a surrogate model. Both methods allow rapid parametrization of polymer field-theoretic models without the need for additional experiments. We evaluate these approaches on data sets involving polymer–solvent–nonsolvent ternary mixtures and block copolymer–solvent systems, demonstrating their robustness to experimental noise and their relevance for real-world formulation design.

copolymers↗
NASA NTRS1997

The Gaseous Content of the Universe at Zeta less than 1.6

Together with graduate student Hsiao-Wen Chen, I have measured and analyzed structural and morphological parameters of 38 galaxies in eight fields for which sensitive measurements of corresponding Ly(alpha) absorption toward background QSOs are available. These measurements are based on Wide Field Planetary Camera 2 (WFPC2) observations obtained with the Hubble Space Telescope (HST) and provide a first look at how the incidence and extent of tenuous gas around galaxies depends on galaxy luminosity, size, and morphological type and on geometry of the impact. The primary result of the analysis is that the amount of gas encountered along the line of sight depends on the galaxy impact parameter and B-band luminosity but does not depend strongly on the galaxy average surface brightness, disk-to-bulge ratio, or redshift. This result confirms and improves upon an anti-correlation between Ly(alpha) absorption equivalent width and galaxy impact parameter found previously. More importantly, this result provides the first quantitative means of relating statistics of faint galaxies to statistics of Ly(alpha) absorption systems. which we plan to exploit to constrain the luminosity function of galaxies beyond the realm of current surveys. Results have been submitted for publication and will greatly improve our statistical conclusions . Together with graduate student Noriaki Yahata. I have measured and classified spectral properties of over 1000 faint galaxies and stars obtained in our low-resolution spectroscopic survey. The goal of this project is two-fold: (1) to exhaustively characterize the spectral properties of all faint galaxies that comprise our current survey, and (2) to gain experience with our measurement and classification code. which ultimately will be used on a data base of 20,000 galaxies to be obtained with the Two-Degree Field (2df) spectrograph at the Anglo-Australian Telescope (AAT). The results will ultimately be used for many goals, but so far we have concentrated on using the results to make a binary classification of the galaxies (i.e. early type versus late type) and to then exploit the density-morphology relationship to obtain a crude density indicator. The primary result of the analysis is that the incidence and extent of tenuous gas around galaxies shows no strong preference for local galaxy environment, at least over the range of densities spanned by the current observations. Along similar lines, two instances of Ly(alpha) absorption lines that arise in groups or clusters were examined. Analysis demonstrates that some can produce corresponding absorption lines and that LY(alpha) absorption lines do not avoid a high-density environment. A new measure of the galaxy-absorber cross-correlation function defines the statistical criterion by which galaxies and absorber pairs are to be matched. I have identified a damped Ly(alpha) absorption system at redshift z equals approximately 0.16, the lowest redshift confirmed to date. The most important results of the analysis are learning that the metal abundances of the absorption system are less than 10 percent of the solar metal abundance and that the absorbing gas is not rotating with the galaxy disk.

Source record↗
NASA NTRS1999

Visual-vestibular integration motion perception reporting

Self-orientation and self/surround-motion perception derive from a multimodal sensory process that integrates information from the eyes, vestibular apparatus, proprioceptive and somatosensory receptors. Results from short and long duration spaceflight investigations indicate that: (1) perceptual and sensorimotor function was disrupted during the initial exposure to microgravity and gradually improved over hours to days (individuals adapt), (2) the presence and/or absence of information from different sensory modalities differentially affected the perception of orientation, self-motion and surround-motion, (3) perceptual and sensorimotor function was initially disrupted upon return to Earth-normal gravity and gradually recovered to preflight levels (individuals readapt), and (4) the longer the exposure to microgravity, the more complete the adaptation, the more profound the postflight disturbances, and the longer the recovery period to preflight levels. While much has been learned about perceptual and sensorimotor reactions and adaptation to microgravity, there is much remaining to be learned about the mechanisms underlying the adaptive changes, and about how intersensory interactions affect perceptual and sensorimotor function during voluntary movements. During space flight, SMS and perceptual disturbances have led to reductions in performance efficiency and sense of well-being. During entry and immediately after landing, such disturbances could have a serious impact on the ability of the commander to land the Orbiter and on the ability of all crew members to egress from the Orbiter, particularly in a non-nominal condition or following extended stays in microgravity. An understanding of spatial orientation and motion perception is essential for developing countermeasures for Space Motion Sickness (SMS) and perceptual disturbances during spaceflight and upon return to Earth. Countermeasures for optimal performance in flight and a successful return to Earth require the development of preflight and in-flight training to help astronauts acquire and maintain a dual adaptive state. Despite the considerable experience with, and use of, an extensive set of countermeasures in the Russian space program, SMS and perceptual disturbances remain an unresolved problem on long-term flights. Reliable, valid perceptual reports are required to develop and refine stimulus rearrangements presented in the PAT devices currently being developed as countermeasures for the prevention of motion sickness and perceptual disturbances during spaceflight, and to ensure a less hazardous return to Earth. Prior to STS-8, crew member descriptions of their perceptual experiences were, at best, anecdotal. Crew members were not schooled in the physiology or psychology of sensory perception, nor were they exposed to the appropriate professional vocabulary. However, beginning with the STS-8 Shuttle flight, a serious effort was initiated to teach astronauts a systematic method to classify and quantify their perceptual responses in space, during entry, and after flight. Understanding, categorizing, and characterizing perceptual responses to spaceflight has been greatly enhanced by implementation of that training system.

Harm, Deborah L.↗
NASA NTRS2016

40 Years of Processing Pieces of Space

This year marks the 40th year anniversary for the Antarctic Search for Meteorite (ANSMET) program. In 1976, the ANSMET program led the first expedition to Antarctica. The ANSMET program is a US-led field-based science project that recovers meteorite samples from Antarctica. Once a year from late November to late January, a field team consisting of 8 to 12 people, spends 6-8 weeks camping on the ice and collecting meteorites. Since 1976, more than 22,000 meteorite samples have been recovered. These meteorites come from asteroids, planets and other bodies of the solar system. Once collected, the Antarctic meteorites are shipped to NASA/Johnson Space Center (JSC) Houston, TX. in a refrigerated truck and are kept frozen to minimize oxidation until they are ready for initial processing. In Antarctica each meteorite is given a field tag which consists of numbers, once in the lab, this is replaced by an official tag, consisting of the Antarctic field location and year collected. The types and numbers of meteorites that have been classified include 849 carbonaceous chondrites, 135 enstatites, 512 achondrites, 64 stony, 115 irons, 48 others (27 R chondrites, 7 ungrouped), 6,161 H chondrites, 7,668 L chondrites, and 4,589 LL chondrites. Although 80-85 percent of the collected meteorites fall in the ordinary chondrite group, the other approximately 15 percent represent rare types of achondrites and carbonaceous chondrites. These rare meteorites include 25 lunar meteorites, 15 Martian meteorites, scores of various types of carbonaceous chondrites, and unique achondrites. The Antarctic meteorites that have been collected are processed in the Meteorite Processing Lab at JSC in Houston, TX. Initial processing of the meteorites begins with thawing/drying the meteorites in a nitrogen glove box for 24 to 48 hours. The meteorites are then photographed, measured, weighed and a description of the interior and exterior of each meteorite is written. The meteorite is broken and a representative sample, either a 1-3 gram chip or thin section is sent to the Smithsonian Institution for classification. After Antarctic meteorites have been classified and approved by the Nomenclature Committee of the Meteoritical Society, they are announced in the Antarctic Meteorite Newsletter, which is published twice per year (fall and spring) so that scientists may review which meteorites are available to study. Requests for Antarctic Meteorite samples are welcomed from research scientists, regardless of their current state of funding for meteorite studies. Since its inception over 3,300 requests have been made for pieces of these meteorites and over 400 investigators worldwide are active in the study of meteorites.. Research on these samples has been published in more than1500 peer reviewed articles; a listing of papers for any meteorite sample can be generated by accessing http://curator.jsc.nasa.gov/antmet/referencesearch.cfm. Antarctic meteorite samples requested by scientists are prepared several different ways. Most samples are prepared as chips, either using a rock splitter or using a chisel and chipping bowl. In special situations, a researcher may request a meteorite slab in which case the samples are cut using a diamond-bladed bandsaw inside of a dry nitrogen glove box. The meteorites are always cut in a 100 percent liquid-free environment. Additionally, thin/thick sections of Antarctic meteorites are also prepared at JSC. The meteorite thin section lab at JSC can prepare standard 30-micron thin sections, thick sections of variable thickness (100 to 200 microns), or demountable sections using superglue, all section are prepared without using water. Although many of the techniques used back in the '70's are still used today, advances in computers, software, databases, available tools and instrumentation have helped to streamline and shorten the duration of the classification process. In conjunction with present day missions to asteroids and other planets, meteorite studies have not only led to a better understanding of the complex histories of these bodies but have also tied certain meteorite groups to particular asteroid bodies. New meteorite discoveries by the ANSMET program provide a cost effective method for obtaining samples of previously unsampled bodies, allowing scientists to learn more about the origin, composition, and evolution of the solar system. Preservation in our cleanrooms at NASA allows material to be archived for future generations and advances in instrumentation and analysis.

Satterwhite, C. E.↗
NASA NTRSDate not supplied

Towards an Aviation Large Language Model by Fine-tuning and Evaluating Transformers

In the aviation domain, there are many applications for machine learning and artificial intelligence tools that utilize natural language. For example, there is a desire to know the commonalities in written safety reports such as voluntary post incidents reports or aerial wildfire operations reports to better understand the risks present. Another use-case is the possibility of extracting airspace procedures and constraints currently written in documents such as Letters of Agreement. These applications can benefit from the use of state-of-the-art natural language processing techniques when adapted to the language/phraseology specific to the aviation domain. This paper evaluates the viability of adaptation of NLP tools to the aviation domain by fine-tuning transformer based models using aviation data sets. In 2018, a novel language model based on neural units (also called transformers) was created and became known as “Bidirectional Encoder Representations from Transformers” or BERT. This architecture combined with large amounts of English training data and innovative semi-supervised training tasks set the standard for what would later emerge as Large Language Models. The performance of these models was further improved by hyperparameter tuning and refinement of the semi-supervised training task and resulted in “Robustly Optimized BERT Pre-training Approach through hyperparameter tuning” or RoBERTa models. These pre-trained Large Language Models proved to be useful for a wide variety of natural language processing tasks such as text classification and question answering through a process called fine-tuning. The transformer architecture with pre-trained weights served as the basis with the last few layers replaced with layers fine-tuned to perform a new task e.g., a layer that provides a label for the entire input text. This process of fine-tuning can also be used to adapt the models to new domains; e.g., BioBERT started with the pre-trained BERT model and was completed by additional fine-tuning and training on biomedical documents. Transformer-based architectures can also be used to create rich representations of text called embeddings which can serve as the input to other machine learning models. This allows simpler algorithms such as logistic regression to use context-rich representations of the text while still remaining quick to train and evaluate. In the world of aviation, there is a growing demand for natural language processing and understanding but the domain presents unique challenges. Due to the technical content (and specialized language) of most aviation documents, fine-tuning pre-trained Large Language Models to specific tasks has not met the benchmark on natural language processing tasks set by simpler models trained from scratch on the data. To address this deficiency, this paper evaluates the improvements from fine-tuning a Large Language Model on a large set of aviation documents using the original semi-supervised training tasks before performing specific natural language tasks. In fine-tuning, a domain-specific dataset is used on the original training task but with the pre-trained Large Language Model instead of starting from a random initialization. This approach allows the model to be adapted to the specific domain language without discarding the information gained from training on general English data. This paper utilized two major dataset types to train and assess the RoBERTa fine-tuning performance. The first are 7,057 Letters of Agreement which are Federal Aviation Administration (FAA) documents that formalize airspace operations across the national airspace system. They contain many examples of ‘aviation English’ using domain specific terminology and phrasing which serves as a representative basis to perform the semi-supervised fine-tuning. The second type is the 494 document classification labels to be used for evaluation. This down-stream evaluation aims to show the performance of the fine-tuned model, better understand how much data is needed for an effective fine-tuning, and how fine-tuning can be adapted for different applications in-the domain. After semi-supervised training, evaluation begins by encoding the documents for classification using the fine-tuned RoBERTa model. Then a logistic regression classifier is trained to label the document type and compared against our ground truth labels. This currently leads to a 82.8% accuracy on 10-fold cross validation showing improvement over baseline RoBERTa which achieved 81.0%. We plan to measure the improvements on additional tasks and it is expected that these improvements will lead to more robust models that can tackle the natural language processing challenges present in aviation datasets.

ATM↗
DOE OSTI2025

Mechanical separations of corn stover anatomical fractions in an integrated feedstock preprocessing system: An experimental and data-driven modeling study

High variabilities of material attributes in lignocellulosic biomass present risks for biofuel and biochemical productions and must be mitigated via preprocessing. Since almost no mechanical device is originally designed for processing biomass, how to operate existing apparatuses with efficient performance has not been investigated extensively. This work presents a study on an integrated screening and air classification to separate cobs and stalks from husks and leaves in corn stover. Prototype machine learning models were developed to assess the feasibility of predicting the process outcome based on the measurable parameters. The models trained upon limited experimental data rendered decent predictive accuracy of yield and purity. The experimental data and modeling results collectively suggest decreasing throughput leads to a higher purity. To the contrary, if throughput increases, a lower purity is likely. A possible trade-off between yield and purity of the separated streams indicates the need for optimal combinations of feedstock size, moisture, and throughput to achieve optimized separations. The results of this study also suggest the need to further improve model predictability by developing more accurate formulations for physics governing the integrated unit operations. To accomplish this, additional experimental data needs to be generated for model training.

09 - BIOMASS FUELS↗
NASA NTRS2025

Navigating Team Dynamics: Automated Detection of Micro-Behaviors Between Team Members Through Longitudinal Interaction Data

The success in future long term space exploration missions will depend on the cooperation, coordination, and mutual understanding among the crew members. Micro-behaviors are momentary, subtle linguistic and paralinguistic indicators of thinking and feeling toward another member of the team (Cortina et al., 2001; Smith & Griffiths, 2022) that can significantly impact team dynamics and influence the overall team performance (Paromita & Chaspari, 2024). Due to their interactive nature, micro-behaviors have a sender (i.e., the team member expressing the micro-behavior) and a target (the team member impacted by the micro-behavior). Detection of these behaviors can assist in avoiding possible conflict among crew members and promoting the overall team success. Our prior research focused on an initial proof of concept of machine learning (ML) models and natural language processing (NLP) techniques that were used for automatically detect micro-behaviors among crew members of the US National Aeronautics and Space Administration’s (NASA) Human Exploration Research Analog (HERA) Campaigns 4 and 5 missions (Paromita et al., 2023). Results underscored the importance of incorporating contextual information in the ML models in the form of sentiment analysis, type of task, and dyadic interaction among team members. Here, we expand the scope of our prior work in two ways. First, we assess ML/NLP methods on new behavioral annotations coded using an adapted version of Smith & Griffins (2022) theoretical framework in terms of Violation (i.e., presence of valenced behavior, uplifting/positive or discouraging/negative), Intensity (i.e., force of behavior in terms of how uplifting or discouraging is the behavior), and Intent (i.e., motive of the behavior in terms of whether it was deliberate or unintentional). Second, we expand the design of the ML model to preserve information about the role of each team member within the occurrence of the micro-behavior (in contrast to the previous model that only considered the sender and the target without determining the team member role). This allows to consider all team members' contributions in the conversation and model long-term dependencies in the dialogue. Our experiments for this study are conducted on data from 5 teams of the NASA HERA C4 (NASA grant NNX16AQ48G (PI: Bell)). Conversations were extracted from the 1.5 hour Team Interaction Battery (TIB) task that occurred 5 times in-mission per crew. This resulted in a total of 13,058 conversational turns (i.e., 17.8% uplifting, 3.3% discouraging, 75.76% neutral, 3.14% nulls). Our findings with the revised behavioral coding and ML/NLP models indicate a 43.66% macro F1-score (i.e., 38.29% precision (P), 50.8% recall (R)) for a dialog state-tracking model that includes information from the sender only, and a 40.9% F1-score (i.e., 38.7% P, 43.36% R) for the same model that includes information from both the sender and the target of the micro-behavior. These are significantly higher compared to simple random forest models that classify behaviors strictly based on speech content and do not consider iterative team dynamics, achieving a 36.07% F1-score (i.e., 39.04% R, 33.53% P). Our findings demonstrate potential ways to leverage large conversational datasets to better capture complex team dynamics. We will discuss future directions including proposed models that can incorporate additional mission days and tasks beyond the TIB for objectively quantifying team behavior at high temporal resolution in space exploration missions.

Projna Paromita↗
DOE OSTI2024

Development of message passing-based graph convolutional networks for classifying cancer pathology reports

Abstract Background Applying graph convolutional networks (GCN) to the classification of free-form natural language texts leveraged by graph-of-words features (TextGCN) was studied and confirmed to be an effective means of describing complex natural language texts. However, the text classification models based on the TextGCN possess weaknesses in terms of memory consumption and model dissemination and distribution. In this paper, we present a fast message passing network (FastMPN), implementing a GCN with message passing architecture that provides versatility and flexibility by allowing trainable node embedding and edge weights, helping the GCN model find the better solution. We applied the FastMPN model to the task of clinical information extraction from cancer pathology reports, extracting the following six properties: main site, subsite, laterality, histology, behavior, and grade. Results We evaluated the clinical task performance of the FastMPN models in terms of micro- and macro-averaged F1 scores. A comparison was performed with the multi-task convolutional neural network (MT-CNN) model. Results show that the FastMPN model is equivalent to or better than the MT-CNN. Conclusions Our implementation revealed that our FastMPN model, which is based on the PyTorch platform, can train a large corpus (667,290 training samples) with 202,373 unique words in less than 3 minutes per epoch using one NVIDIA V100 hardware accelerator. Our experiments demonstrated that using this implementation, the clinical task performance scores of information extraction related to tumors from cancer pathology reports were highly competitive.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2024

Organ-delimited gene regulatory networks provide high accuracy in candidate transcription factor selection across diverse processes

Organ-specific gene expression datasets that include hundreds to thousands of experiments allow the reconstruction of organ-level gene regulatory networks (GRNs). However, creating such datasets is greatly hampered by the requirements of extensive and tedious manual curation. Here, we trained a supervised classification model that can accurately classify the organ-of-origin for a plant transcriptome. This K-Nearest Neighbor-based multiclass classifier was used to create organ-specific gene expression datasets for the leaf, root, shoot, flower, and seed in Arabidopsis thaliana . A GRN inference approach was used to determine the: i. influential transcription factors (TFs) in each organ and, ii. most influential TFs for specific biological processes in that organ. These genome-wide, organ-delimited GRNs (OD-GRNs), recalled many known regulators of organ development and processes operating in those organs. Importantly, many previously unknown TF regulators were uncovered as potential regulators of these processes. As a proof-of-concept, we focused on experimentally validating the predicted TF regulators of lipid biosynthesis in seeds, an important food and biofuel trait. Of the top 20 predicted TFs, eight are known regulators of seed oil content, e.g., WRI1, LEC1, FUS3. Importantly, we validated our prediction of MybS2, TGA4, SPL12, AGL18, and DiV2 as regulators of seed lipid biosynthesis. We elucidated the molecular mechanism of MybS2 and show that it induces purple acid phosphatase family genes and lipid synthesis genes to enhance seed lipid content. This general approach has the potential to be extended to any species with sufficiently large gene expression datasets to find unique regulators of any trait-of-interest.

09 BIOMASS FUELS↗
NASA NTRSDate not supplied

A Machine Learning Approach to Improve Air Traffic Management Initiatives

Collaborating closely with commercial air carriers and related organizations, the Federal Aviation Administration(FAA) regulates air traffic and ensures the safety and efficiency of air operations. Air traffic controllers make strategic decisions, such as delaying, rerouting, or canceling flights, partly based on guidance provided by the FAA’s Air TrafficControl System Command Center (ATCSCC). The guidance includes, among other things, control measures known asTraffic Management Initiatives (TMIs) designed to enhance safety and improve operational efficiency. TMIs play a crucial role in managing the demand and capacity within the U.S. National Airspace System (NAS). Two major TMIs that are routinely used (primarily to mitigate the adverse effects of bad weather) are Ground Delay Programs (GDPs) andGround Stops (GSs). In a GDP, flights destined for airports facing thunderstorm activity experience delays at their origin airports. This proactive approach minimizes the risk of routing aircraft through hazardous weather conditions and also replaces (fuel burning) airborne delays with ground delays. In a GS, a temporary restriction is imposed on the departure or arrival of aircraft at a specific airport or within a designated airspace. Although other TMIs (e.g., miles-in-trail) are also implemented as part of (air) traffic flow management in the NAS, the focus of this work is on GDPs and GSs. Since TMIs, by design, lead to flight delays or cancellations, it is crucial to put in place the right set of parameters(e.g., scope and duration of the GDP). For example, when the end time of a GDP extends beyond what is necessary, it imposes unnecessary delays on departing flights. This situation could occur as a result of inaccurate prediction of the(required) duration of the GDP based on the weather forecast. On the other hand, if a GDP ends prematurely before the underlying capacity constraints are resolved at the destination airport, it may result in airborne holding. The delicate balance lies in matching the termination of the GDP precisely with the resolution of capacity constraints, avoiding both the imposition of unnecessary ground delays and the need for airborne holding due to premature program termination.Failing to specify the right parameters for TMIs also leads to flight delays, creating a significant obstacle in managing the increasing traffic volumes causing increased work load for the controllers. To address this issue, we propose the integration of Machine Learning (ML) models in the traffic flow management(TFM) pipeline. In current operations, decisions are made by human experts based on extensive training, historical patterns, available traffic and weather data. Since we have an abundance of data from past events that tell us the likely impact of various TMIs, by ingesting historical data, properly trained ML models can offer valuable insights and aid human decision-making. With the FAA increasingly exploring advanced analytics, ML emerges as a focal point for enhancing TFM within the National Airspace System (NAS). As a first step, this study aims to provide traffic controllers with decision-making support for the issuance and adjustment of TMIs. Data analytics and machine learning have been previously employed to address some of the challenges associated with TMIs. Numerous studies have concentrated on various facets of TMI issuance, exploring factors influencing TMI parameters, including arrival rate, airport capacity, and delay prediction. For example, using weather forecasts, several statistical methods were used to produce probabilistic capacity profiles which in conjunction with deterministic models provided insights into the GDP planning process [1–4]. The downside of using deterministic models is that they rely on fixed inputs and predetermined rules, which lack the ability to account for the inherent uncertainty and variability present in real-world scenarios. In a separate series of studies, researchers aimed to predict the occurrences of GDPs and GSs. The majority of these studies utilized various supervised learning methods, including Decision Trees, Naive Bayes, Support VectorMachines, and Random Forests to analyze the influence of weather conditions and arrival demand on TMI incidents[5–8]. However, these studies primarily focused on predicting the incidence of TMIs without explicitly addressing the scope of TMIs, including their duration and their geographical coverage. Furthermore, the emphasis of these studies was largely on GDPs, given their higher frequency and longer duration when compared to GSs. A limited number of studies focused on predicting the parameters of TMIs, specifically addressing their duration and extent. In one such study focusing on optimizing the TMI parameters at San Francisco International Airport (SFO),the authors utilized a probabilistic forecast of fog [9]. They simulated various capacity scenarios based on the (fog)burn-off forecasts, selecting GDP parameters that minimized airborne and overall ground delays. However, this approach exclusively emphasizes stratus (fog) burn-off as the primary determinant of GDP and GS, neglecting other influential factors like severe weather events, runway closures, lower capacity than traffic demand, and other important variables. Given the complexity of predicting the TMI and determining its scope, we seek a more holistic approach. We aim to consider all significant factors that could impact TMIs and their parameters. What sets this research apart is the fusion of all data sources relevant to the issuance and adjustment of TMIs and it represents the first comprehensive attempt to optimize TMIs in this manner. Since this comprehensive solution involves various aspects, we break down the problem into smaller components and input all parameters into a unified model called the “TMI Adjuster”. Figure 1 shows the overall framework and the list of datasets used in each model. The objective of the TMI Adjuster module is to deliver reliable, consistent and expedited recommendations for the progression, adjustment, and termination of TMIs. The ML solution entails developing a pipeline capable of predicting the necessity of a TMI (e.g., GS or GDP) along with its various parameters. For example, in the case of a GS, this includes the scope of the GS either in terms of distance from the destination airport or based on pre-defined airspace sectors. Here, scope refers to those regions and departing airports that are subject to the GS. In this paper, we concentrate on the issuance of GSs in the three major airports in the New York area — LaGuardia(LGA), John F. Kennedy International (JFK), and Newark Liberty International (EWR). We fuse traffic, weather and other relevant aviation data from years 2017 to 2019 to train and validate the ML models. In particular, we use the following datasets: •Terminal Aerodrome Forecast (TAF): meteorological forecasts specific to each airport, issued four times a day, covering predefined time periods. •TMI data: includes all GSs and GDPs along with their respective parameters. •Aviation System Performance Metrics (ASPM): includes traffic related data such as aircraft delays, arrival, and departure rates. •Notices to Airmen (NOTAMs): utilized to extract runway closure data and manage interdependencies between terminals in close proximity. •Flight cancellation data •Airspace Flow Programs (AFP): includes information on flight airborne holdings caused by TMIs. The data preprocessing entails transforming ASPM, TMI, AFP, NOTAMs, and weather data into an hourly format and consolidating all datasets by merging them based on date and time as the primary key. The TMI Adjuster framework comprises two parallel models: one dedicated to GS and a second model focused on GDP. As previously mentioned, our specific focus is on the GS model as a multi-classification problem. In this framework, each data point of the GS model input summarizes ten hours of data. Specifically, the data loader for the GS model generates the input and output of the model as follows: at a given time step, the input includes the actual traffic, weather, and TMI data from the two-hour window before the time step, alongside the weather forecast and scheduled traffic for the next 8 hours starting from the time step. Based on this information, the output of the GS model for each time interval consists of three dimensions. The first dimension represents a binary decision on whether there should be a GS in place for the next hour or not. The second dimension is related to the scope of the GS in the United States, and the third dimension is related to the scope of the GS in Canada (i.e., to determine if the GS impacts airports in Canada).One of the challenges with TMI modeling is the sparsity of TMI events, particularly regarding its scope. To address this challenge in the scope of the GS model output, we implement grouping. The GS scope for the US region is defined based on a list of centers that should be included when the GS is in place. With 20 centers in the US, we utilized historical data to group them into 4 categories. In particular, we summarized our historical data in a graph format where nodes represent centers, and link weights are defined based on the co-occurrence of centers in the scope parameter ofTMIs. By identified strongly connected components in this graph, we were able to partition the centers into four groups. We consider two model structures for the GS Model. Firstly, a hierarchical classification model [10], where the human decision-making for a GS is of hierarchical nature. The decision-maker first decides whether there is a need fora GS, and if the answer is yes, determines the scope. A hierarchical classification model organizes the problem into a class hierarchy, typically a tree or a Directed Acyclic Graph (DAG) structure, and considers the dependency of the decision in the previous step to the next component [10]. Here, we employ the local classifier per level approach, which involves training one multi-class classifier for each level of the class hierarchy. The second structure is the independent structure. In this setting, as the name suggests, we do not consider the dependency of the decisions in the different dimensions of the output of the model. Instead, for each dimension, we train a multi-class classifier independently. Table 1 summarizes GS model statistics for training, validation and testing. The table documents the effect of limiting data to the time steps when there was actually a TMI in place or when a TMI had just terminated. This resulted in a more balanced distribution of the GS class(GS positive class)versus “No GS”(GS negative class), which might help the training process. While JFK and LGA follow very similar distributions, with 40% and 42% GS positive class respectively, EWR has proportionally fewer GS incidents at 28%. Our subsequent phase involves evaluating the performance of both hierarchical structure and independent structure using different state-of-the-art multi-class classifier models such as Random Forest, Decision Trees, K-nearest Neighbors, and Logistic Regression and forecast the duration and scope of the GSs.

Farzan Masrour Shalmani↗