Search NASA⌕ Search

SEARCH · Search NASA

Results for “pattern classification”

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 505 records · Page 28

Gamma-Ray Burst Class Properties

Guided by the Supervised pattern recognition algorithm C4.5, we examine the three gamma-ray burst classes identified by Mukherjee et al. C4.5 provides strong statistical support for this classification. However, with C4.5 and our knowledge of the BATSE instrument, we demonstrate that Class 3 (intermediate fluence, intermediate duration, soft) does not have to be a distinct source population: statistical/systematic errors in measuring burst attributes combined with the well-known hardness/intensity correlation can cause low peak flux Class I (high fluence, long, intermediate hardness) bursts to take on Class 3 characteristics naturally. Based on our hypothesis that the third class is not a distinct one, we provide rules so that future events can be placed in either Class I or Class 2 (low fluence, short, hard). Using classified bursts from the BATSE 4B Catalog, we plot log(N>P) vs. log(P) curves and study spectral features of each class. We find that the two classes are relatively distinct on the basis of spectral parameters, alpha, Beta, and E(sub peak) alone. Although this does not indicate a better basis for classification, it does suggest that different physical conditions exist for Class I and Class 2 bursts. In the process of studying burst class characteristics, we identify a new bias that affects measurement of burst fluences and durations. Using a simple model of how burst duration can be underestimated, we generally characterize how this fluence duration bias affects BATSE measurements, and demonstrate the type of effect it can have on the BATSE fluence vs. peak flux diagram.

Hakkila, Jon↗

Detection of anomalous events

A system is described for receiving a stream of events and scoring the events based on anomalousness and maliciousness (or other classification). The system can include a plurality of anomaly detectors that together implement an algorithm to identify low-probability events and detect atypical traffic patterns. The anomaly detector provides for comparability of disparate sources of data (e.g., network flow data and firewall logs.) Additionally, the anomaly detector allows for regulatability, meaning that the algorithm can be user configurable to adjust a number of false alerts. The anomaly detector can be used for a variety of probability density functions, including normal Gaussian distributions, irregular distributions, as well as functions associated with continuous or discrete variables.

97 MATHEMATICS AND COMPUTING↗

Fast and interpretable classification of small X-ray diffraction datasets using data augmentation and deep neural networks

X-ray diffraction (XRD) data acquisition and analysis is among the most time-consuming steps in the development cycle of novel thin-film materials. We propose a machine-learning-enabled approach to predict crystallographic dimensionality and space group from a limited number of thin-film XRD patterns. We overcome the scarce-data problem intrinsic to novel materials development by coupling a supervised machine learning approach with a model-agnostic, physics-informed data augmentation strategy using simulated data from the Inorganic Crystal Structure Database (ICSD) and experimental data. As a test case, 115 thin-film metal-halides spanning 3 dimensionalities and 7 space-groups are synthesized and classified. After testing various algorithms, we develop and implement an all convolutional neural network, with cross-validated accuracies for dimensionality and space-group classification of 93% and 89%, respectively. We propose average class activation maps, computed from a global average pooling layer, to allow high model interpretability by human experimentalists, elucidating the root-causes of misclassification. Finally, we systematically evaluate the maximum XRD pattern step size (data acquisition rate) before loss of predictive accuracy occurs, and determine it to be 0.16° 2θ, which enables an XRD pattern to be obtained and classified in 5.5 minutes or less.

36 MATERIALS SCIENCE↗

Unsupervised classification for region of interest in X-ray ptychography

X-ray ptychography offers high-resolution imaging of large areas at a high computational cost due to the large volume of data provided. To address the cost issue, we propose a physics-informed unsupervised classification algorithm that is performed prior to reconstruction and removes data outside the region of interest (RoI) based on the multimodal features present in the diffraction patterns. The preprocessing time for the proposed method is inconsequential in contrast to the resource-intensive reconstruction process, leading to an impressive reduction in the data workload to a mere 20% of the initial dataset. This capability consequently reduces computational time dramatically while preserving reconstruction quality. Through further segmentation of the diffraction patterns, our proposed approach can also detect features that are smaller than beam size and correctly classify them as within the RoI.

97 MATHEMATICS AND COMPUTING↗

Interpretable machine learning models classify minerals via spectroscopy

Developing methods to identify mineral species confidently and rapidly from Raman spectral analysis is critical to numerous fields. Traditionally, analysis relies on pattern matching the Raman spectrum of an unknown dataset with a supporting library of well-characterized spectral data, which may prove difficult for environmental samples that are poorly crystalline or phase mixtures. Here, we developed interpretable machine learning models that can classify uranium minerals by secondary oxyanion chemistry and other physicochemical properties based solely on Raman spectra. This new ML method produces a mineral profile of physical and chemical properties for an unknown sample and can rapidly classify or identify unknown minerals from Raman data, without the need for an exact pattern match in a spectral library. Training models are validated by 1. Strong correlation of high confidence model regions with published spectroscopic assignments and 2. Correct classification of a mineral not present in training data. Training data are from the Compendium of Uranium Raman and Infrared Experimental Spectra and available crystallographic information files within the open-source Smart Spectral Matching scientific framework. Physically meaningful classifier models can rapidly identify key structural and chemical information about unknown uranium minerals and the overall methodology is broadly applicable for mineral phases.

Machine learning↗

On the consistency of Monte Carlo track structure DNA damage simulations

Purpose: Monte Carlo track structures (MCTS) simulations have been recognized as useful tools for radiobiological modeling. However, the authors noticed several issues regarding the consistency of reported data. Therefore, in this work, they analyze the impact of various user defined parameters on simulated direct DNA damage yields. In addition, they draw attention to discrepancies in published literature in DNA strand break (SB) yields and selected methodologies. Methods: The MCTS code Geant4-DNA was used to compare radial dose profiles in a nanometer-scale region of interest (ROI) for photon sources of varying sizes and energies. Then, electron tracks of 0.28 keV–220 keV were superimposed on a geometric DNA model composed of 2.7 × 10{sup 6} nucleosomes, and SBs were simulated according to four definitions based on energy deposits or energy transfers in DNA strand targets compared to a threshold energy E{sub TH}. The SB frequencies and complexities in nucleosomes as a function of incident electron energies were obtained. SBs were classified into higher order clusters such as single and double strand breaks (SSBs and DSBs) based on inter-SB distances and on the number of affected strands. Results: Comparisons of different nonuniform dose distributions lacking charged particle equilibrium may lead to erroneous conclusions regarding the effect of energy on relative biological effectiveness. The energy transfer-based SB definitions give similar SB yields as the one based on energy deposit when E{sub TH} ≈ 10.79 eV, but deviate significantly for higher E{sub TH} values. Between 30 and 40 nucleosomes/Gy show at least one SB in the ROI. The number of nucleosomes that present a complex damage pattern of more than 2 SBs and the degree of complexity of the damage in these nucleosomes diminish as the incident electron energy increases. DNA damage classification into SSB and DSB is highly dependent on the definitions of these higher order structures and their implementations. The authors’ show that, for the four studied models, different yields are expected by up to 54% for SSBs and by up to 32% for DSBs, as a function of the incident electrons energy and of the models being compared. Conclusions: MCTS simulations allow to compare direct DNA damage types and complexities induced by ionizing radiation. However, simulation results depend to a large degree on user-defined parameters, definitions, and algorithms such as: DNA model, dose distribution, SB definition, and the DNA damage clustering algorithm. These interdependencies should be well controlled during the simulations and explicitly reported when comparing results to experiments or calculations.

60 APPLIED LIFE SCIENCES↗

OpenCRUMS USA: An Open Machine Learning Framework for Characterizing Variability in Aerosol Reanalysis Data

Advances in artificial intelligence (AI) have called for exploring how these techniques can be used for exploring patterns in large climate datasets. To that regard, the U.S. Department of Energy AI for Earth System Predictability (AI4ESP) supported a pilot initiative called the Open Classification of Regimes in the Southeast USA (OpenCRUMS USA) project to explore how AI can be used to characterize modes of spatial variability in large climate datasets. For this study, we focus on comparing two methods for characterizing the modes of spatial variability of surface aerosol concentration over the Houston region: empirical orthogonal functions (EOFs) and layerwise relevance propagation (LRP) applied to a convolutional neural network (CNN) classifier. We show that EOF analysis typically attributes spatial variability modes that span all of southeast Texas, prohibiting the attribution of spatial variability to localized regions. However, using LRP on the CNN classifier resolves the explanatory parameters at a finer spatial resolution than EOFs. This allows for the attribution of the spatial variability of surface aerosols to local regions of organic carbon which was not possible using EOFs. In addition, the LRP analysis also suggests that synoptic-scale transport of dust is most prevalent during anticyclonic and pretrough synoptic conditions as categorized by self-organizing maps.

54 ENVIRONMENTAL SCIENCES↗

Feasibility study for locating archaeological village sites by satellite remote sensing techniques

The author has identified the following significant results. The objective is to determine the feasibility of detecting large Alaskan archaeological sites by satellite remote sensing techniques and mapping such sites. The approach used is to develop digital multispectral signatures of dominant surface features including vegetation, exposed soils and rock, hydrological patterns and known archaeological sites. ERTS-1 scenes are then printed out digitally in a map-like array with a letter reflecting the most appropriate classification representing each pixel. Preliminary signatures were developed and tested. It was determined that there was a need to tighten up the archaeological site signature by developing accurate signatures for all naturally-occurring vegetation and surface conditions in the vicinity of the test area. These second generation signatures have been tested by means of computer printouts and classified tape displays on the University of Alaska CDU-200 and by comparison with aerial photography. It has been concluded that the archaeological signatures now in use are as good as can be developed. Plans are to print out signatures for the entire test area and locate on topographic maps the likely locations of archaeological sites within the test area.

Cook, J. P.↗

Performance analysis of image processing algorithms for classification of natural vegetation in the mountains of southern California

The earth's forests fix carbon from the atmosphere during photosynthesis. Scientists are concerned that massive forest removals may promote an increase in atmospheric carbon dioxide, with possible global warming and related environmental effects. Space-based remote sensing may enable the production of accurate world forest maps needed to examine this concern objectively. To test the limits of remote sensing for large-area forest mapping, we use Landsat data acquired over a site in the forested mountains of southern California to examine the relative capacities of a variety of popular image processing algorithms to discriminate different forest types. Results indicate that certain algorithms are best suited to forest classification. Differences in performance between the algorithms tested appear related to variations in their sensitivities to spectral variations caused by background reflectance, differential illumination, and spatial pattern by species. Results emphasize the complexity between the land-cover regime, remotely sensed data and the algorithms used to process these data.

Yool, S. R.↗

The Value of Long-Term (40 years) Airborne Gamma Radiation SWE Record for Evaluating Three Observation-Based Gridded SWE Data Sets by Seasonal Snow and Land Cover Classifications

Observation‐based long‐term gridded snow water equivalent (SWE) products are important assets for hydrological and climate research. However, an evaluation of the currently available SWE products has been limited due to the lack of independent SWE data that extend over a large range of environmental conditions. In this study, three daily long‐term SWE products (Special Sensor Microwave Imager and Sounder [SSMI/S] SWE, GlobSnow‐2 SWE, and University of Arizona [UA] SWE) we reevaluated by seasonal snow cover and land cover classifications over the conterminous United States from1982 to 2017, using the historical airborne gamma radiation SWE observations (20,738 measurements).We found that there are similar patterns in SSMI/S and GlobSnow‐2 SWE when compared against the gamma SWE. However, GlobSnow‐2 SWE had better agreement with gamma SWE than SSMI/S SWE in some forested‐type classes and maritime and prairie snow classes. As compared to SSMI/S and GlobSnow‐2SWE, UA SWE has much better agreement with gamma SWE in all land cover types and snow classes. Tree cover and topographic heterogeneity affect the agreement between the gamma and gridded SWE and accuracy of gamma SWE itself with the largest differences typically occurring when the percent tree cover was 80% or higher, the terrain slope was steeper than 2.5°, and the elevation range exceeded 100 m. The results demonstrate the reliability of the UA SWE products and the benefits of the gamma radiation approach to measure SWE, especially in forested regions.

Eunsang Cho↗

Spatiotemporal Methane Emissions from Global Lakes and Reservoirs

Inland aquatic systems, such as lakes and reservoirs, contribute substantially to global methane (CH4) emissions; yet are among the most uncertain components of the total CH4 budget. Lakes and reservoirs have received recent attention as they may generate high CH4 fluxes. Improved quantification of these CH4 fluxes, particularly their spatiotemporal distribution, is key to realistically incorporating them in CH4 modeling and budget studies. Here we report on a new global, gridded (0.25° lat × 0.25° lon) study of lake and reservoir CH4 emissions, accounting for new knowledge regarding lake and reservoir areal extent and distribution, and spatiotemporal emission patterns influenced by diurnal variability, temperature-dependent seasonality, satellite-derived freeze-thaw dynamics, and eco-climatic and physical CH4-centric type classification. The results of this new data set comprise daily CH4 emissions from lake and reservoirs throughout the full annual cycle and are tightly anchored to field observations, in situ measurements, and remote-sensing observations. Results show that reservoirs cover 297 × 103 km2 globally and emit 10.1 Tg CH4 yr-1 from diffusive (1.2 Tg CH4 yr-1) and ebullitive (8.9 Tg CH4 yr-1) emission pathways. On a global scale, CH4 emitting areas of lakes cover 1853 × 103 km2 and emit 37.1 Tg CH4 yr-1 from diffusive (17.1 Tg CH4 yr-1) and ebullitive (23.0 Tg CH4 yr-1) emission pathways and an additional ebullition flux of 5.2 Tg CH4 yr-1 upon ice-melt due to the accumulation of bubbles during the freeze period. This analysis of lakes and reservoir CH4 emission addresses multiple gaps and uncertainties in previous studies and represents an important contribution to studies of the global CH4 budget. The new data sets and methodologies from this study provide a framework to better understand and model the current and future role of lakes and reservoirs in the global CH4 budget and to guide efforts to mitigate inland aquatic system CH4 emissions. This presentation will describe the methodologies applied and the major results of this study focusing on the spatiotemporal distribution of lake and reservoir type classification, processes driving lake and reservoir emission seasonality, and the contribution of individual lake and reservoir types to the annual cycle of inland aquatic CH4 emissions.

Spatiotemporal↗

Chemical classification program synthesis using generative artificial intelligence

Accurately classifying chemical structures is essential for cheminformatics and bioinformatics, including tasks such as identifying bioactive compounds of interest, screening molecules for toxicity to humans, finding non-organic compounds with desirable material properties, or organizing large chemical libraries for drug discovery or environmental monitoring. However, manual classification is labor-intensive and difficult to scale to large chemical databases. Existing automated approaches either rely on manually constructed classification rules, or are deep learning methods that lack explainability. This work presents an approach that uses generative artificial intelligence to automatically write chemical classifier programs for classes in the Chemical Entities of Biological Interest (ChEBI) database. These programs can be used for efficient deterministic run-time classification of SMILES structures, with natural language explanations. The programs themselves constitute an explainable computable ontological model of chemical class nomenclature, which we call the ChEBI Chemical Class Program Ontology (C3PO). We validated our approach against the ChEBI database, and compared our results against deep learning models and a naive SMARTS pattern based classifier. C3PO outperforms the naive classifier, but does not reach the performance of state of the art deep learning methods. However, C3PO has a number of strengths that complement deep learning methods, including explainability and reduced data dependence. C3PO can be used alongside deep learning classifiers to provide an explanation of the classification, where both methods agree. The programs can be used as part of the ontology development process, and iteratively refined by expert human curators.

Artificial Intelligence↗

The Pixel Anomaly Detection Tool : a user-friendly GUI for classifying detector frames using machine-learning approaches

Data collection at X-ray free electron lasers has particular experimental challenges, such as continuous sample delivery or the use of novel ultrafast high-dynamic-range gain-switching X-ray detectors. This can result in a multitude of data artefacts, which can be detrimental to accurately determining structure-factor amplitudes for serial crystallography or single-particle imaging experiments. Here, a new data-classification tool is reported that offers a variety of machine-learning algorithms to sort data trained either on manual data sorting by the user or by profile fitting the intensity distribution on the detector based on the experiment. This is integrated into an easy-to-use graphical user interface, specifically designed to support the detectors, file formats and software available at most X-ray free electron laser facilities. The highly modular design makes the tool easily expandable to comply with other X-ray sources and detectors, and the supervised learning approach enables even the novice user to sort data containing unwanted artefacts or perform routine data-analysis tasks such as hit finding during an experiment, without needing to write code.

47 OTHER INSTRUMENTATION↗

Hierarchical Mixture of Experts for Advanced Air Mobility Flight Phase Classification

Advanced Air Mobility (AAM) and Urban Air Mobility (UAM) operations will have numerous vehicles and aircraft flying in the airspace, which poses safety and security concerns. Commercial airlines utilize Air Traffic Management (ATM) and Air Traffic Control (ATC) for real-time monitoring, surveillance, traffic coordination, and rerouting to maintain safe and efficient flight patterns. Transferring ATM and ATC architectures to AAM/UAM will be challenging to implement since AAM/UAM aircraft fly at lower altitudes, have more static and dynamic obstacles, operate in highly dense environments, and have several more aircraft to monitor for a given volume of the national airspace (NAS). Aircraft typically have the following flight phases: takeoff, climb, cruise, descent, and landing. Classifying these flight phases provides insight into ensuring safe operations, providing situational awareness of the NAS, and monitoring flights in emergencies. Automatic flight phase classification will enhance the efficiencies of ATM/ATC-like architectures for AAM/UAM, especially since numerous aircraft will be flying in highly dense urban environments. Typical flight phase classification methods are all-or-nothing, which will not capture or accurately classify the transitions between flight phases. Utilizing hierarchical mixture of experts (HME) provides a flight phase classification solution that includes transitions between the flight phases by assigning weights based on ground-based distributed sensor readings from cameras and radar. Adding the transitions between flight phases increases the fidelity of flight phase classification and provides deeper insight into flight phase classification by leveraging distributed sensing concepts. Simulation results and post-processed flight test results demonstrate the utility of HME for automatic and robust flight phase classification for real-time AAM operations.

distributed sensing↗

TPSAS-NF1676L-32014-DND

The Cloud-Aerosol Lidar with Orthogonal Polarization (CALIOP), on-board the Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observations (CALIPSO) is a satellite-borne polarization sensitive lidar. It has been providing the vertical distributions of clouds and aerosols along with their microphysical and optical properties since 2006. One of its important Level 2 products, feature classification, has been determined using the lidar information from 532 nm parallel and perpendicular channels, and 1064 nm channel measurements of layer integrated backscatter. Deep machine learning methods which combine both the channel and texture information to recognize feature patterns is uniquely beneficial when applied to this data. In this study, we will use Convolutional Neural Network (CNN), a deep machine learning method, to classify lidar aerosol subtypes by using the lidar profile observations. This method uses additional information from the vertical texture of the feature instead of using only the layer information. Note that in the integrated layer properties, the texture information has been masked due to averaging. Our results will show how the texture information plays a role in the classification. This preliminary work explores the benefits and potential of deep machine learning methods for lidar retrievals and focuses on the aerosol subtype classification. The broader application extends to the classification of other feature types. Future applications include the developing deep machine learning methods with neural networks to retrieve properties of the features, and studies of indirect effect of cloud-aerosol interaction from lidar measurements.

Shan Zeng Kowalski↗

Optical recognition of statistical patterns

Optical implementation of the Fukunaga-Koontz transform (FKT) and the Least-Squares Linear Mapping Technique (LSLMT) is described. The FKT is a linear transformation which performs image feature extraction for a two-class image classification problem. The LSLMT performs a transform from large dimensional feature space to small dimensional decision space for separating multiple image classes by maximizing the interclass differences while minimizing the intraclass variations. The FKT and the LSLMT were optically implemented by utilizing a coded phase optical processor. The transform was used for classifying birds and fish. After the F-K basis functions were calculated, those most useful for classification were incorporated into a computer generated hologram. The output of the optical processor, consisting of the squared magnitude of the F-K coefficients, was detected by a T.V. camera, digitized, and fed into a micro-computer for classification. A simple linear classifier based on only two F-K coefficients was able to separate the images into two classes, indicating that the F-K transform had chosen good features. Two advantages of optically implementing the FKT and LSLMT are parallel and real time processing.

Lee, S. H.↗

A knowledge-informed large language model framework for U.S. nuclear power plant shutdown initiating event classification for probabilistic risk assessment

Identifying and classifying shutdown initiating events (SDIEs) is critical for developing shutdown probabilistic risk assessment for nuclear power plants. Existing computational approaches cannot achieve satisfactory performance due to the challenges of unavailable large, labeled datasets, imbalanced event types, and label noise. To address these challenges, we propose a hybrid pipeline that integrates a knowledge-informed machine learning model to prescreen non-SDIEs and a large language model (LLM) to classify SDIEs into four types. In the prescreening stage, we proposed a set of 44 SDIE text patterns that consist of the most salient keywords and phrases from six SDIE types. Text vectorization based on the SDIE patterns generates feature vectors that are highly separable by using a simple binary classifier. The second stage builds Bidirectional Encoder Representations from Transformers (BERT)-based LLM, which learns generic English language representations from self-supervised pretraining on a large dataset and adapts to SDIE classification by fine-tuning it on an SDIE dataset. The proposed approaches are evaluated on a dataset with 10,928 events using precision, recall ratio, F 1 score, and average accuracy. In conclusion, the results demonstrate that the prescreening stage can exclude more than 97% non-SDIEs, and the LLM achieves an average accuracy of 95.1% for SDIE classification.

99 - GENERAL AND MISCELLANEOUS↗

Decimated Input Ensembles for Improved Generalization

Recently, many researchers have demonstrated that using classifier ensembles (e.g., averaging the outputs of multiple classifiers before reaching a classification decision) leads to improved performance for many difficult generalization problems. However, in many domains there are serious impediments to such "turnkey" classification accuracy improvements. Most notable among these is the deleterious effect of highly correlated classifiers on the ensemble performance. One particular solution to this problem is generating "new" training sets by sampling the original one. However, with finite number of patterns, this causes a reduction in the training patterns each classifier sees, often resulting in considerably worsened generalization performance (particularly for high dimensional data domains) for each individual classifier. Generally, this drop in the accuracy of the individual classifier performance more than offsets any potential gains due to combining, unless diversity among classifiers is actively promoted. In this work, we introduce a method that: (1) reduces the correlation among the classifiers; (2) reduces the dimensionality of the data, thus lessening the impact of the 'curse of dimensionality'; and (3) improves the classification performance of the ensemble.

Tumer, Kagan↗