SEARCH · Search NASA

Results for “Classification Problem”

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.

SourceResearch connectionFrom yearThrough year

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 307 records · Page 17

All sources NASA NTRS DOE OSTI

DOE OSTI2020

Updates to Relevance Vector Machine: Multiclass Classification, Variable Selection, and Proof-of-Concept Application to Safeguards Fresh Fuel Verification using List-Mode Neutron Collar Data

To expand the capabilities of safeguards authorities to verify the integrity of fresh fuel assemblies, Oak Ridge National Laboratory has retrofit the existing electronics of the JCC-71 uranium neutron coincidence collar, which contains 18 3 He neutron detectors and an external 241 AmLi(α, n) neutron interrogation source arranged to surround a fresh nuclear fuel assembly. The new electronics system allows analysts to record list-mode neutron multiplicity data in addition to the singles and doubles rates that are currently measured. Based on previous proof-of-concept research, analysis of these new data will identify off-normal fuel configurations in an assembly and characterize or localize the specific partial fuel defects. The purpose of this report it to document the analysis algorithm development and then to demonstrate its capability for the safeguards verification of fresh fuel assemblies using list mode neutron collar data. To analyze the complex list-mode data collected with the upgraded uranium neutron collar, multivariate classification algorithms are being developed using a novel classification method, the relevance vector machine. This approach may be applied to multiclass problems to estimate the probability that test data belongs to one of many possible classes of data. In addition, our method identifies the most useful variables/channels for making predictions, which illuminates the basis for the model’s predictions, and this interpretability is largely unique among data analytics methods. Variable selection occurs during model training and parameter tuning and does not need any external hyperparameter tuning routines. Finally, we apply the modified relevance vector machine to a simulated dataset of list-mode neutron collar data generated with the radiation transport code MCNP. The method can correctly identify off-normal fuel configurations, categorize the data according to four fuel defect scenarios, and rank the channels in the data according to prediction utility. For nuclear safeguards applications, it is concluded that this method has the potential to increase the sensitivity and reliability to detect missing fuel rods from a standard 17 x 17 Pressurized Water Reactor (PWR) fresh fuel assembly. Within this analysis, “off-normal” (i.e., missing fuel rods) were correctly classified in 17 simulated test scenarios with one quarter (25%) of the fresh fuel rods missing using a training data set of 58 simulated measurements.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗

DOE OSTI2023

Predicting Building Envelope Construction from In-Situ Thermal Testing

When embarking on a retrofit of a building envelope, it is critical to understand the composition of its assembly. This practice is currently done by destructive and invasive material testing or demolition, which is sometimes not possible when in historic or protected buildings. To address this problem, in-situ thermal testing can be utilized along with machine learning classification algorithms to infer the composition of an assembly. In this paper, a proof-of-concept K-nearest neighbors classification model is developed to classify assembly composition from effective thermal resistance, effective thermal mass, and assembly cladding. This model was trained and tested utilizing a synthetic dataset producing an F1-score of 94.6%. This model was also validated with experimental data from a 100-year old wall assembly, confirming the model’s real-world validity. The paper presents a framework for inferring as-built envelope assemblies, all without having to damage or disturb the building and its occupants.

Pilet, Tyler J.↗

DOE OSTI2023

Densely Connected G-invariant Deep Neural Networks with Signed Permutation Representations

We introduce and investigate, for finite groups G, G-invariant deep neural network (GDNN) architectures with ReLU activation that are densely connected- i.e., include all possible skip connections. In contrast to other G-invariant architectures in the literature, the preactivations of theG-DNNs presented here are able to transform by signed permutation representations (signed perm-reps) of G. Moreover, the individual layers of the G-DNNs are not required to be G-equivariant; instead, the preactivations are constrained to be G-equivariant functions of the network input in a way that couples weights across all layers. The result is a richer family of G-invariant architectures never seen previously. We derive an efficient implementation of G-DNNs after a reparameterization of weights, as well as necessary and sufficient conditions for an architecture to be "admissible"- i.e., nondegenerate and inequivalent to smaller architectures. We include code that allows a user to build a G-DNN interactively layer-by-layer, with the final architecture guaranteed to be admissible. We show that there are far more admissible G-DNN architectures than those accessible with the "concatenated ReLU" activation function from the literature. Finally, we apply G-DNNs to two example problems--(1) multiplication in --1, 1} (with theoretical guarantees) and (2) 3D object classification--finding that the inclusion of signed perm-reps significantly boosts predictive performance compared to baselines with only ordinary (i.e., unsigned) perm-reps.

97 MATHEMATICS AND COMPUTING↗

NASA NTRS1975

Maximum likelihood signature estimation

Maximum-likelihood estimates are discussed which are based on an unlabeled sample of observations, of unknown parameters in a mixture of normal distributions. Several successive approximation procedures for obtaining such maximum-likelihood estimates are described. These procedures, which are theoretically justified by the local contractibility of certain maps, are designed to take advantage of good initial estimates of the unknown parameters. They can be applied to the signature extension problem, in which good initial estimates of the unknown parameters are obtained from segments which are geographically near the segments from which the unlabeled samples are taken. Additional problems to which these methods are applicable include: estimation of proportions and adaptive classification (estimation of mean signatures and covariances).

Walker, H. F.↗

NASA NTRS1988

Space Station power system autonomy demonstration

The Systems Autonomy Demonstration Program (SADP) represents NASA's major effort to demonstrate, through a series of complex ground experiments, the application and benefits of applying advanced automation technologies to the Space Station project. Lewis Research Center (LeRC) and Marshall Space Flight Center (MSFC) will first jointly develop an autonomous power system using existing Space Station testbed facilities at each center. The subsequent 1990 power-thermal demonstration will then involve the cooperative operation of the LeRC/MSFC power system with the Johnson Space Center (JSC's) thermal control and DMS/OMS testbed facilities. The testbeds and expert systems at each of the NASA centers will be interconnected via communication links. The appropriate knowledge-based technology will be developed for each testbed and applied to problems requiring intersystem cooperation. Primary emphasis will be focused on failure detection and classification, system reconfiguration, planning and scheduling of electrical power resources, and integration of knowledge-based and conventional control system software into the design and operation of Space Station testbeds.

Kish, James A.↗

NASA NTRS1998

Smart Optical RAM for Fast Information Management and Analysis

Statement of Problem Instruments for high speed and high capacity in-situ data identification, classification and storage capabilities are needed by NASA for the information management and analysis of extremely large volume of data sets in future space exploration, space habitation and utilization, in addition to the various missions to planet-earth programs. Parameters such as communication delays, limited resources, and inaccessibility of human manipulation require more intelligent, compact, low power, and light weight information management and data storage techniques. New and innovative algorithms and architecture using photonics will enable us to meet these challenges. The technology has applications for other government and public agencies.

Liu, Hua-Kuang↗

NASA NTRS2014

A Classification Table for Achondrites

Classifying chondrites is relatively easy and the criteria are well documented. It is based on mineral compositions, textural characteristics and more recently, magnetic susceptibility. It can be more difficult to classify achondrites, especially those that are very similar to terrestrial igneous rocks, because mineralogical, textural and compositional properties can be quite variable. Achondrites contain essentially olivine, pyroxenes, plagioclases, oxides, sulphides and accessory minerals. Their origin is attributed to differentiated parents bodies: large asteroids (Vesta); planets (Mars); a satellite (the Moon); and numerous asteroids of unknown size. In most cases, achondrites are not eye witnessed falls and some do not have fusion crust. Because of the mineralogical and magnetic susceptibility similarity with terrestrial igneous rocks for some achondrites, it can be difficult for classifiers to confirm their extra-terrestrial origin. We -as classifiers of meteorites- are confronted with this problem with every suspected achondrite we receive for identification. We are developing a "grid" of classification to provide an easier approach for initial classification. We use simple but reproducible criteria based on mineralogical, petrological and geochemical studies. We presented the classes: acapulcoites, lodranites, winonaites and Martian meteorites (shergottite, chassignites, nakhlites). In this work we are completing the classification table by including the groups: angrites, aubrites, brachinites, ureilites, HED (howardites, eucrites, and diogenites), lunar meteorites, pallasites and mesosiderites. Iron meteorites are not presented in this abstract.

Chennaoui-Aoudjehane, H.↗

NASA NTRS1974

Classification, scale series and nomenclature of lunar charts

Lunar charts are classified with an eye toward solution of special problems of lunar cartography. The initial state of mapping the moon will require charts with scales of 1:5,000,000; 1:1,000,000; 1:250,000; 1:50,000; and 1:10,000.

Florenskiy, K. P.↗

NASA NTRS2016

Statistical Emulator for Expensive Classification Simulators

Expensive simulators prevent any kind of meaningful analysis to be performed on the phenomena they model. To get around this problem the concept of using a statistical emulator as a surrogate representation of the simulator was introduced in the 1980's. Presently, simulators have become more and more complex and as a result running a single example on these simulators is very expensive and can take days to weeks or even months. Many new techniques have been introduced, termed criteria, which sequentially select the next best (most informative to the emulator) point that should be run on the simulator. These criteria methods allow for the creation of an emulator with only a small number of simulator runs. We follow and extend this framework to expensive classification simulators.

Ross, Jerret↗

DOE OSTI2020

Benders Cut Classification via Support Vector Machines for Solving Two-Stage Stochastic Programs

In this work, we consider Benders decomposition for solving two-stage stochastic programs with complete recourse based on finite samples of the uncertain parameters. We define the Benders cuts binding at the final optimal solution or the ones significantly improving bounds over iterations as valuable cuts. We propose a learning-enhanced Benders decomposition (LearnBD) algorithm, which adds a cut classification step in each iteration to selectively generate cuts that are more likely to be valuable cuts. The LearnBD algorithm includes two phases: (i) sampling cuts and collecting information from training problems and (ii) solving testing problems with a support vector machine (SVM) cut classifier. We run the LearnBD algorithm on instances of capacitated facility location and multicommodity network design under uncertain demand. Our results show that SVM cut classifier works effectively for identifying valuable cuts, and the LearnBD algorithm reduces the total solving time of all instances for different problems with various sizes and complexities.

97 MATHEMATICS AND COMPUTING↗

NASA NTRS1992

Magnetic Repeat Station Data

The nature, purpose, availability and distribution of magnetic repeat stations data are reviewed. Attention is paid to how repeat station data are obtained, sources of errors and limitations on accuracy. Repeat station results cannot achieve the accuracy of magnetic observatory estimates of the secular variation, but are, nevertheless, essential for improving the limited spatial coverage provided by observatories. The lack of appropriate standards and reporting procedures have meant that most global field molders have made little, or no, use of repeat data. Partly to rectify this problem, IAGA Working Group V-4 has instigated a scheme for systematic reporting and classification of magnetic repeat station data. The scheme will help to ensure that proper use is made of the wealth of repeat data that is available, and will also promote better observational practices.

Barton, C. E.↗

NASA NTRS1993

Classification of simple vegetation types using POLSAR image data

Mapping basic vegetation or land cover types is a fairly common problem in remote sensing. Knowledge of the land cover type is a key input to algorithms which estimate geophysical parameters, such as soil moisture, surface roughness, leaf area index or biomass from remotely sensed data. In an earlier paper, an algorithm for fitting a simple three-component scattering model to POLSAR data was presented. The algorithm yielded estimates for surface scatter, double-bounce scatter and volume scatter for each pixel in a POLSAR image data set. In this paper, we show how the relative levels of each of the three components can be used as inputs to simple classifier for vegetation type. Vegetation classes include no vegetation cover (e.g. bare soil or desert), low vegetation cover (e.g. grassland), moderate vegetation cover (e.g. fully developed crops), forest and urban areas. Implementation of the approach requires estimates for the three components from all three frequencies available using the NASA/JPL AIRSAR, i.e. C-, L- and P-bands. The research described in this paper was carried out by the Jet Propulsion Laboratory, California Institute of Technology under a contract with the National Aeronautics and Space Administration.

Freeman, A.↗

NASA NTRS1994

Neural network application to comprehensive engine diagnostics

We have previously reported on the use of neural networks for detection and identification of faults in complex microprocessor controlled powertrain systems. The data analyzed in those studies consisted of the full spectrum of signals passing between the engine and the real-time microprocessor controller. The specific task of the classification system was to classify system operation as nominal or abnormal and to identify the fault present. The primary concern in earlier work was the identification of faults, in sensors or actuators in the powertrain system as it was exercised over its full operating range. The use of data from a variety of sources, each contributing some potentially useful information to the classification task, is commonly referred to as sensor fusion and typifies the type of problems successfully addressed using neural networks. In this work we explore the application of neural networks to a different diagnostic problem, the diagnosis of faults in newly manufactured engines and the utility of neural networks for process control.

Marko, Kenneth A.↗

NASA NTRS1993

Classification of Simple Vegetation Types Using POLSAR Image Data

Freeman, A.↗

NASA NTRS1986

Incipient fault detection study for advanced spacecraft systems

A feasibility study to investigate the application of vibration monitoring to the rotating machinery of planned NASA advanced spacecraft components is described. Factors investigated include: (1) special problems associated with small, high RPM machines; (2) application across multiple component types; (3) microgravity; (4) multiple fault types; (5) eight different analysis techniques including signature analysis, high frequency demodulation, cepstrum, clustering, amplitude analysis, and pattern recognition are compared; and (6) small sample statistical analysis is used to compare performance by computation of probability of detection and false alarm for an ensemble of repeated baseline and faulted tests. Both detection and classification performance are quantified. Vibration monitoring is shown to be an effective means of detecting the most important problem types for small, high RPM fans and pumps typical of those planned for the advanced spacecraft. A preliminary monitoring system design and implementation plan is presented.

Milner, G. Martin↗

NASA NTRS1976

An ad hoc map evaluation procedure

An ad hoc map evaluation procedure is proposed which is most suitable for evaluating low-resolution classification maps against high resolution ground truth maps, such as maps against interpreted aircraft photographs. Commonly practiced sampling and evaluation procedures are impracticable in this context because of difficulties in registration and in comparing the samples. This ad hoc procedure is designed to overcome these two major problems, and its practicability is discussed. Two widely accepted parameters are estimated by the new procedure; namely, the probability of correct classification and the proportion biases. Statistical qualifications are also provided.

Kan, E. P.↗

NASA NTRS1987

Gaussian maximum likelihood and contextual classification algorithms for multicrop classification

The paper reviews some of the ways in which context has been handled in the remote-sensing literature, and additional possibilities are introduced. The problem of computing exhaustive and normalized class-membership probabilities from the likelihoods provided by the Gaussian maximum likelihood classifier (to be used as initial probability estimates to start relaxation) is discussed. An efficient implementation of probabilistic relaxation is proposed, suiting the needs of actual remote-sensing applications. A modified fuzzy-relaxation algorithm using generalized operations between fuzzy sets is presented. Combined use of the two relaxation algorithms is proposed to exploit context in multispectral classification of remotely sensed data. Results on both one artificially created image and one MSS data set are reported.

Di Zenzo, Silvano↗

NASA NTRS1981

Comet nucleus - Some characteristics and a hypothesis on origin and structure

The spectroscopic properties of comets are discussed, and consideration is given to the problems of the cosmic ray irradiation and the origin and structure of cometary nuclei. The classification of cometary spectra as continuum dominant, molecular emissions dominant and CO(+) dominant is examined, and changes in spectral characteristics with heliocentric distance although not with cometary age are pointed out. Spectral evidence for a variety of organic compounds in the cometary nucleus is noted, and a scenario is presented whereby these complex molecules may form as a result of cosmic ray irradiation of cometary ices in the Oort cloud. Finally, a model is proposed for the accumulation of small icy grains to form a cometary nucleus composed of cometesimals of size distribution following a -5th power dependence on mass.

Donn, B.↗