Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Classification bias”

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 145 records · Page 8
All sourcesNASA NTRSDOE OSTI
DOE OSTI2021

Uranium Oxide Synthetic Pathway Discernment through Unsupervised Morphological Analysis

We present a novel unsupervised machine learning method for quantitative representation of scanning electron micrographs and its applications and performance for nuclear forensic analysis of uranium ore concentrates. The method uses a vector quantizing variational autoencoder followed by a histogram operation to encode a micrograph into a single dimensional representation, called the latent vector. The method requires no extant labeling of the data and can be applied over large datasets of micrographs with minimal human interaction. The representations generated are broadly descriptive of each micrograph and the microstructure of the material imaged. In the case of uranium ore concentrate analysis, the representations were amenable to processing reagent and ore concentrate species classification with accuracy of 81:8%, which is competitive with state-of-the-art supervised networks. The representations were also used to classify previously unseen processing routes, were able to classify imaging parameters such as magnification (to 76:0% accuracy), were able to classify fine grained process parameters such as calcining temperature (to 74:4% accuracy), and their informatic properties indicate that they are generally descriptive of the image represented. This method can be applied across microstructure analysis fields to perform quantitative analysis without the need for labor intensive and possibly biased human analysis.

Scanning Electron Microscopy, Vector Quantizing Va↗
NASA NTRS1976

Evaluation of algorithms for estimating wheat acreage from multispectral scanner data

The author has identified the following significant results. Fourteen different classification algorithms were tested for their ability to estimate the proportion of wheat in an area. For some algorithms, accuracy of classification in field centers was observed. The data base consisted of ground truth and LANDSAT data from 55 sections (1 x 1 mile) from five LACIE intensive test sites in Kansas and Texas. Signatures obtained from training fields selected at random from the ground truth were generally representative of the data distribution patterns. LIMMIX, an algorithm that chooses a pure signature when the data point is close enough to a signature mean and otherwise chooses the best mixture of a pair of signatures, reduced the average absolute error to 6.1% and the bias to 1.0%. QRULE run with a null test achieved a similar reduction.

Nalepka, R. F.↗
NASA NTRS2009

Astrometry With the Hubble Space Telescope: Trigonometric Parallaxes of Planetary Nebula Nuclei NGC 6853, NGC 7293, ABELL 31, and DeHt 5

We present absolute parallaxes and relative proper motions for the central stars of the planetary nebulae NGC 6853 (The Dumbbell), NGC 7293 (The Helix), Abell 31, and DeHt 5. This paper details our reduction and analysis using DeHt 5 as an example. We obtain these planetary nebula nuclei (PNNi) parallaxes with astrometric data from Fine Guidance Sensors FGS 1r and FGS 3, white-light interferometers on the Hubble Space Telescope. Proper motions, spectral classifications and VJHKT2M and DDO51 photometry of the stars comprising the astrometric reference frames provide spectrophotometric estimates of reference star absolute parallaxes. Introducing these into our model as observations with error, we determine absolute parallaxes for each PNN. Weighted averaging with previous independent parallax measurements yields an average parallax precision, sigma (sub pi)/ pi = 5%. Derived distances are: d(sub NGC6853) = 405(exp +28 sub -25) pc, d(sub NGC7293) = 216(exp +14 sub -12) pc, d(sub Abell31) = 621(exp +91 sub -70) pc, and d(sub DeHt5) = 345(exp +19 sub -17) pc. These PNNi distances are all smaller than previously derived from spectroscopic analyses of the central stars. To obtain absolute magnitudes from these distances requires estimates of interstellar extinction. We average extinction measurements culled from the literature, from reddening based on PNNi intrinsic colors derived from model SEDs, and an assumption that each PNN experiences the same rate of extinction as a function of distance as do the reference stars nearest (in angular separation) to each central star. We also apply Lutz-Kelker bias corrections. The absolute magnitudes and effective temperatures permit estimates of PNNi radii through both the Stefan-Boltzmann relation and Eddington fluxes. Comparing absolute magnitudes with post-AGB models provides mass estimates. Masses cluster around 0.57 solar Mass, close to the peak of the white dwarf mass distribution. Adding a few more PNNi with well-determined distances and masses, we compare all the PNNi with cooler white dwarfs of similar mass, and confirm, as expected, that PNNi have larger radii than white dwarfs that have reached their final cooling tracks.

STARS↗
NASA NTRS1982

Profile modeling for crop discrimination

A complete automatic unsupervised classification approach has been developed to estimate the proportion of corn, soybeans, and other crops in a Landsat 5-by-6-nanometer segment. The technique involves extracting crop growth variables that are predictable from agromet models, it provides a signature extension not possible before. The technique has been applied in 3 years (1978-1980) for 56 segments and shows neither a year-to-year effect nor any geographic effect. The results to date are considered far better than those obtainable from any other currently available technique. Nontheless, the technique can be improved to reduce the bias.

Badhwar, G. D.↗
DOE OSTI2020

A Template-based Approach to the Photometric Classification of SN 1991bg-like Supernovae in the SDSS-II Supernova Survey

The use of SNe Ia to measure cosmological parameters has grown significantly over the past two decades. However, there exists a significant diversity in the SN Ia population that is not well understood. Overluminous SN 1991T-like and subluminous SN 1991bg-like objects are two characteristic examples of peculiar SNe. The identification and classification of such objects is an important step in studying what makes them unique from the remaining SN population. With the upcoming Vera C. Rubin Observatory promising on the order of a million new SNe over a 10 year survey, spectroscopic classifications will be possible for only a small subset of observed targets. As such, photometric classification has become an increasingly important concern in preparing for the next generation of astronomical surveys. Using observations from the Sloan Digital Sky Survey II (SDSS-II) SN Survey, we apply here an empirically based classification technique targeted at the identification of SN 1991bg-like SNe in photometric data sets. By performing dedicated fits to photometric data in the rest-frame redder and bluer bandpasses, we classify 16 previously unidentified 91bg-like SNe. Using SDSS-II host galaxy measurements, we find that these SNe are preferentially found in host galaxies with an older average stellar age than the hosts of normal SNe Ia. We also find that these SNe are found at a further physical distance from the center of their host galaxies. We find no statistically significant bias in host galaxy mass or specific star formation rate for these targets.

Astronomy & Astrophysics↗
NASA NTRS2018

Boundary of the Slow Solar Wind

This work argues that there are two fundamental states of the non-transient solar wind, and that these can be distinguished by a number of criteria. Here we define the states, which will be termed slow and fast, or SSW and FSW, for lack of better terms, by the level of velocity fluctuations, v, in them, with the slow wind having systematically lower fluctuations than the fast wind. Almost all winds with speeds less than 450 km s1 are in the slow class, and winds with speeds greater than 600 km s1 are fast, but we argue that in between, consistent with other work, the v classification is more fundamental than speed. We show that the fluctuation categorization coincides well with classes based on Alfvénicy, proton specific entropy, ion thermal speed, and ionic composition. This correlated behavior among these solar wind parameters exists regardless of it being associated with a heliospheric current sheet or a pseudostreamer. This work provides evidence that both the so-called SSW I and SSW II scenarios coexist for the SSW formation. In addition, that the dynamical properties (thermal, magnetic, and turbulence properties) correlate well with properties set at the inner corona (ion ionization states and FIP bias)implies that there exists a boundary layer on the Sun within which the SSW is formed. This boundary layer would set up the coronal conditions for the source and transport of the SSW.

Ko, Yuan-Kuen↗
NASA NTRS1980

Preliminary evaluation of spectral, normal and meteorological crop stage estimation approaches

Several of the projects in the AgRISTARS program require crop phenology information, including classification, acreage and yield estimation, and detection of episodal events. This study evaluates several crop calendar estimation techniques for their potential use in the program. The techniques, although generic in approach, were developed and tested on spring wheat data collected in 1978. There are three basic approaches to crop stage estimation: historical averages for an area (normal crop calendars), agrometeorological modeling of known crop-weather relationships agrometeorological (agromet) crop calendars, and interpretation of spectral signatures (spectral crop calendars). In all, 10 combinations of planting and biostage estimation models were evaluated. Dates of stage occurrence are estimated with biases between -4 and +4 days while root mean square errors range from 10 to 15 days. Results are inconclusive as to the superiority of any of the models and further evaluation of the models with the 1979 data set is recommended.

Cate, R. B.↗
NASA NTRSDate not supplied

From Simulation to Reality With Random Noise

The challenging environment of autonomous vehicle (AV) navigation necessitates certain functions be performed by deep neural networks. Optimizing these models involves collecting vast quantities of domain-specific training data and ensuring that the dataset is representative of expected conditions. High-fidelity simulation plays a vital role in making this process feasible, allowing a wide range of scenarios to be explored at low cost. However, learning from simulation introduces subtle biases into models, which can degrade real-world performance in unpredictable ways. This effect can be mitigated with learning schemes specialized to bridge distributional shifts (transfer learning). Given the complex nature of these methods, the underlying models, and their environments, meaningfully evaluating performance is notstraight forward. Many unrelated factors can effect an improvement in generalization accuracy, but a full ablation analysis is often difficult. To tease out signal from noise, it is necessary to understand how transfer learning performance is affected by noise itself. The goals of this paper are (i) to establish a domain randomization baseline for a simple classification transfer learning task and (ii) to validate the RRAV testbed as a platform for further research in sim-to-real learning. We generate imagery from a simulation of NASA Ames Research Center and train a small convolutional neural network (ConvNet) to classify position relative to a centerline. Further models are trained with different types of noise progressively added to the data. The models are deployed aboard the on-site test vehicle to test real-world performance. In our experiments, we find that such naive domain randomization raises sim-to-real accuracy from 64% to 79%, while training directly on real data yields an 89% accuracy ceiling. These results suggest that the isolated mechanism of domain randomization can significantly improve generalization.

simulation↗
NASA NTRS1990

Dumb-bell galaxies in southern clusters: Catalog and preliminary statistical results

The dominant galaxy of a rich cluster is often an object whose formation and evolution is closely connected to the dynamics of the cluster itself. Hoessel (1980) and Schneider et al. (1983) estimate that 50 percent of the dominant galaxies are either of the dumb-bell type or have companions at projected distances less than 20 kpc, which is far in excess of the number expected from chance projection (see also Rood and Leir 1979). Presently there is no complete sample of these objects, with the exception of the listing of dumb-bell galaxies in BM type I and I-II clusters in the Abell statistical sample of Rood and Leir (1979). Recent dynamical studies of dumb-bell galaxies in clusters (Valentijn and Casertano, 1988) still suffer from inhomogeneity of the sample. The fact that it is a mixture of optically and radio selected objects may have introduced an unknown biases, for instance if the probability of radio emission is enhanced by the presence of close companions (Stocke, 1978, Heckman et al. 1985, Vettolani and Gregorini 1988) a bias could be present in their velocity distribution. However, this situation is bound to improve: a new sample of Abell clusters in the Southern Hemisphere has been constructed (Abell et al., 1988 hereafter ACO), which has several advantages over the original northern catalog. The plate material (IIIaJ plates) is of better quality and reaches fainter magnitudes. This makes it possible to classify the cluster types with a higher degree of accuracy, as well as to fainter magnitudes. The authors therefore decided to reconsider the whole problem constructing a new sample of dumb-bell galaxies homogeneously selected from the ACO survey. Details of the classification criteria are given.

Vettolani, G.↗
NASA NTRS2001

Linear and Order Statistics Combiners for Pattern Classification

Several researchers have experimentally shown that substantial improvements can be obtained in difficult pattern recognition problems by combining or integrating the outputs of multiple classifiers. This chapter provides an analytical framework to quantify the improvements in classification results due to combining. The results apply to both linear combiners and order statistics combiners. We first show that to a first order approximation, the error rate obtained over and above the Bayes error rate, is directly proportional to the variance of the actual decision boundaries around the Bayes optimum boundary. Combining classifiers in output space reduces this variance, and hence reduces the 'added' error. If N unbiased classifiers are combined by simple averaging. the added error rate can be reduced by a factor of N if the individual errors in approximating the decision boundaries are uncorrelated. Expressions are then derived for linear combiners which are biased or correlated, and the effect of output correlations on ensemble performance is quantified. For order statistics based non-linear combiners, we derive expressions that indicate how much the median, the maximum and in general the i-th order statistic can improve classifier performance. The analysis presented here facilitates the understanding of the relationships among error rates, classifier boundary distributions, and combining in output space. Experimental results on several public domain data sets are provided to illustrate the benefits of combining and to support the analytical results.

Tumer, Kagan↗
NASA NTRS1992

Signal processing for airborne doppler radar detection of hazardous wind shear as applied to NASA 1991 radar flight experiment data

Radar data collected during the 1991 NASA flight tests have been selectively analyzed to support research directed at developing both improved as well as new algorithms for detecting hazardous low-altitude windshear. Analysis of aircraft attitude data from several flights indicated that platform stability bandwidths were small compared to the data rate bandwidths which should support an assumption that radar returns can be treated as short time stationary. Various approaches at detection of weather returns in the presence of ground clutter are being investigated. Non-coventional clutter rejection through spectrum mode tracking and classification algorithms is a subject of continuing research. Based upon autoregressive modeling of the radar return time sequence, this approach may offer an alternative to overcome errors in conventional pulse-pair estimates. Adaptive filtering is being evaluated as a means of rejecting clutter with emphasis on low signal-to-clutter ratio situations, particularly in the presence of discrete clutter interference. An analysis of out-of-range clutter returns is included to illustrate effects of ground clutter interference due to range aliasing for aircraft on final approach. Data are presented to indicate how aircraft groundspeed might be corrected from the radar data as well as point to an observed problem of groundspeed estimate bias variation with radar antenna scan angle. A description of how recorded clutter return data are mixed with simulated weather returns is included. This enables the researcher to run controlled experiments to test signal processing algorithms. In the summary research efforts involving improved modelling of radar ground clutter returns and a Bayesian approach at hazard factor estimation are mentioned.

Baxa, Ernest G., Jr.↗
DOE OSTI2025

Event Classifications on DNE2 Main Experiment Data using a Convolutional Neural Network Ensemble

The Dynamic Networks (DN) Experiment for FY24 (DNE2) is an experiment within DN with the goal of quantitatively evaluating the effectiveness of solutions developed so far by various researchers under the Low Yield Nuclear Monitoring (LYNM) program using a shared set of metrics and datasets. A key component of this experiment is the mimicking of a signature processing pipeline, and comparing currently accepted and standard-use processing methods to more state-of-the-art processes developed under DN. In this work, we focus specifically on the Event Characterization (EC) Focus Area (FA) of the pipeline, where a seismic event’s magnitude, yield and class are identified. We use Deep Learning (DL) to classify the type of events being processed as either earthquakes (EQs) or explosions (EXs) for three iterations of experiment datasets. The model is noticeably more confident and accurate in classifying explosions than earthquakes, reflecting a known shortcoming of the model, that being of a bias towards predicting explosions over earthquakes in the west coast due to training data biases.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2022

Group-Invariant Quantum Machine Learning

Quantum machine learning (QML) models are aimed at learning from data encoded in quantum states. Recently, it has been shown that models with little to no inductive biases (i.e., with no assumptions about the problem embedded in the model) are likely to have trainability and generalization issues, especially for large problem sizes. As such, it is fundamental to develop schemes that encode as much information as available about the problem at hand. In this work we present a simple, yet powerful, framework where the underlying invariances in the data are used to build QML models that, by construction, respect those symmetries. These so-called group-invariant models produce outputs that remain invariant under the action of any element of the symmetry group $\mathfrak{G}$ associated with the dataset. We present theoretical results underpinning the design of $\mathfrak{G}$-invariant models, and exemplify their application through several paradigmatic QML classification tasks, including cases when $\mathfrak{G}$ is a continuous Lie group and also when it is a discrete symmetry group. Notably, our framework allows us to recover, in an elegant way, several well-known algorithms for the literature, as well as to discover new ones. Taken together, we expect that our results will help pave the way towards a more geometric and group-theoretic approach to QML model design.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2021

Rapid and flexible segmentation of electron microscopy data using few-shot machine learning

Abstract Automatic segmentation of key microstructural features in atomic-scale electron microscope images is critical to improved understanding of structure–property relationships in many important materials and chemical systems. However, the present paradigm involves time-intensive manual analysis that is inherently biased, error-prone, and unable to accommodate the large volumes of data produced by modern instrumentation. While more automated approaches have been proposed, many are not robust to a high variety of data, and do not generalize well to diverse microstructural features and material systems. Here, we present a flexible, semi-supervised few-shot machine learning approach for segmentation of scanning transmission electron microscopy images of three oxide material systems: (1) epitaxial heterostructures of SrTiO 3 /Ge, (2) La 0.8 Sr 0.2 FeO 3 thin films, and (3) MoO 3 nanoparticles. We demonstrate that the few-shot learning method is more robust against noise, more reconfigurable, and requires less data than conventional image analysis methods. This approach can enable rapid image classification and microstructural feature mapping needed for emerging high-throughput characterization and autonomous microscope platforms.

36 MATERIALS SCIENCE↗
DOE OSTI2024

Jet classification using high-level features from anatomy of top jets

Recent advancements in deep learning models have significantly enhanced jet classification performance by analyzing low-level features (LLFs). However, this approach often leads to less interpretable models, emphasizing the need to understand the decision-making process and to identify the high-level features (HLFs) crucial for explaining jet classification. To address this, we consider the top jet tagging problems and introduce an analysis model (AM) that analyzes selected HLFs designed to capture important features of top jets. Our AM mainly consists of the following three modules: a relation network analyzing two-point energy correlations, mathematical morphology and Minkowski functionals for generalizing jet constituent multiplicities, and a recursive neural network analyzing subjet constituent multiplicity to enhance sensitivity to subjet color charges. We demonstrate that our AM achieves performance comparable to the Particle Transformer (ParT) while requiring fewer computational resources in a comparison of top jet tagging using jets simulated at the hadronic calorimeter angular resolution scale. Furthermore, as a more constrained architecture than ParT, the AM exhibits smaller training uncertainties because of the bias-variance tradeoff. We also compare the information content of AM and ParT by decorrelating the features already learned by AM. Lastly, we briefly comment on the results of AM with finer angular resolution inputs.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2025

Robustness of Deep Learning Classification to Adversarial Input on GPUs: Asynchronous Parallel Accumulation Is a Source of Vulnerability

The ability of machine learning (ML) classification models to resist small, targeted input perturbations—known as adversarial attacks—is a key measure of their safety and reliability. We show that floating-point non associativity (FPNA) coupled with asynchronous parallel programming on GPUs is sufficient to result in misclassification, without any perturbation to the input. Additionally, we show that this misclassification is particularly significant for inputs close to the decision boundary and that standard adversarial robustness results may be overestimated up to 4.6 when not considering machine-level details. We first study a linear classifier, before focusing on standard Graph Neural Network (GNN) architectures and datasets used in robustness assessments. We develop a novel black-box attack using Bayesian optimization to discover external workloads that can change the instruction scheduling which bias the output of reductions on GPUs and reliably lead to misclassification. Motivated by these results, we present a new learnable permutation (LP) gradient-based approach to learning floating-point operation orderings that lead to misclassifications. The LP approach provides a worst-case estimate in a computationally efficient manner, avoiding the need to run identical experiments tens of thousands of times over a potentially large set of possible GPU states or architectures. Finally, using instrumentation-based testing, we investigate parallel reduction ordering across different GPU architectures under external background workloads, when utilizing multi-GPU virtualization, and when applying power capping. Our results demonstrate that parallel reduction ordering varies significantly across architectures under the first two conditions, substantially increasing the search space required to fully test the effects of this parallel scheduler-based vulnerability. These results and the methods developed here can help to include machine-level considerations into adversarial robustness assessments, which can make a difference in safety and mission critical applications.

Shanmugavelu, Sanjif [Maxeler Technologies, a Groq↗
NASA NTRS2010

High-Resolution Subtropical Summer Precipitation Derived from Dynamical Downscaling of the NCEP-DOE Reanalysis: How Much Small-Scale Information Is Added by a Regional Model?

This study assesses the regional-scale summer precipitation produced by the dynamical downscaling of analyzed large-scale fields. The main goal of this study is to investigate how much the regional model adds smaller scale precipitation information that the large-scale fields do not resolve. The modeling region for this study covers the southeastern United States (Florida, Georgia, Alabama, South Carolina, and North Carolina) where the summer climate is subtropical in nature, with a heavy influence of regional-scale convection. The coarse resolution (2.5deg latitude/longitude) large-scale atmospheric variables from the National Center for Environmental Prediction (NCEP)/DOE reanalysis (R2) are downscaled using the NCEP Environmental Climate Prediction Center regional spectral model (RSM) to produce precipitation at 20 km resolution for 16 summer seasons (19902005). The RSM produces realistic details in the regional summer precipitation at 20 km resolution. Compared to R2, the RSM-produced monthly precipitation shows better agreement with observations. There is a reduced wet bias and a more realistic spatial pattern of the precipitation climatology compared with the interpolated R2 values. The root mean square errors of the monthly R2 precipitation are reduced over 93 (1,697) of all the grid points in the five states (1,821). The temporal correlation also improves over 92 (1,675) of all grid points such that the domain-averaged correlation increases from 0.38 (R2) to 0.55 (RSM). The RSM accurately reproduces the first two observed eigenmodes, compared with the R2 product for which the second mode is not properly reproduced. The spatial patterns for wet versus dry summer years are also successfully simulated in RSM. For shorter time scales, the RSM resolves heavy rainfall events and their frequency better than R2. Correlation and categorical classification (above/near/below average) for the monthly frequency of heavy precipitation days is also significantly improved by the RSM.

NARR↗
DOE OSTI2024

Evaluation of WSR-88D Level III and MRMS Rainfall Estimates Against Rain Gauge Observations at the Savannah River Site

Rainfall data for the Savannah River Site (SRS) has been historically measured by rain gauges. These instruments serve as ground truth for most climatological and weather applications; however, gauge measurements are prone to errors or biases under certain weather conditions. Rainfall estimates from radar reflectivity values have been developed and improved over the years and serve as an alternative or supplement for gauge measurements. This study compares measurements from tipping bucket rain gauges with rainfall estimates from the NOAA NWS WSR-88D Level III hourly rainfall product and the Multi-Radar Multi-Sensor (MRMS) gauge-corrected precipitation estimate. Results show good agreement between radar derived amounts and ground measurements, with MRMS values showing correlation coefficients of 0.60-0.95 and RMSE values of less than 1.0 cm (0.4 in). The WSR-88D Level III estimates result in correlation coefficients of 0.46-0.86 and RMSE values of less than 1.3 cm (0.5 in). Few outliers are observed for each data pair and are evaluated against precipitation classification products (WSR-88D Hybrid Hydrometeor Classification product and MRMS Precipitation Flag product). The rain gauges used in this study are not part of the Hydrometeorological Automated Data System (HADS) network used to correct the MRMS estimates and therefore, results of this work provide an independent validation of the MRMS gauge correction scheme.

54 ENVIRONMENTAL SCIENCES↗