Search NASA⌕ Search

SEARCH · Search NASA

Results for “data 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 703 records · Page 39

A Taxonomic Classification Approach for Global Spatio-temporal Data

The World Bank, World Health Organization, and other major vendors collectively provide thousands of global time series datasets that focus on issues of the environment, public health, economics, violence, education, and national security. Sorting these data into meaningful information requires the use of data mining techniques to cluster trends into an orderly and manageable number of cases. The World SpatioTemporal Analytics and Mapping (WSTAMP) project database (wstamp.ornl.gov) was developed to spatiotemporally harmonize global vendor data (23,300+ attributes, 200+ locations, 50+ years). Within the WSTAMP analytical environment, Dynamic Time Warping (DTW) has been a highly effective data-driven approach for clustering and mapping these time series into national spatiotemporal behavior maps. Two significant properties have surfaced from this work. First, several recognizable cluster patterns have emerged and persist across a range of locations, attributes, and time frames (e.g., increasing, decreasing, rebounding, peak, oscillating). Secondly, practitioners engaging WSTAMP have noted the explanatory and anticipatory value of these patterns and articulated particular interest in detecting them within the spatiotemporal cube. This need was addressed by shifting DTW-based clustering from an open ended, data-driven implementation to a taxonomic pattern matching approach. This paper presents the method including implementation strategies for visualization and human computer interaction and applies the approach to a sample data set and concludes with next steps.

Stewart, Robert↗

Imbalanced Multi-layer Cloud Classification with Advanced Baseline Imager (ABI) and CloudSat/CALIPSO Data

Clouds at different altitudes play different roles in Earth’s climate. Comprehensive understanding of overlapping clouds is important for climate and weather prediction. The East Pacific region is where El Ni˜no and La Ni˜na originate and where multi-layer clouds frequently occur. The overlap of clouds at different altitudes in this region increases the classification complexity for cloud-based climatological studies. Unlike prior work in cloud layer classification that assumes single layer or two-layer of clouds, in this work, we consider multi-layer cloud classification with 8 cloud-level classes (clear-sky, high, middle, low, high+middle, high+low, middle+low, high+middle+low). We develop and analyze machine learning models on features extracted from satellite images from the East Pacific regions collected by GOES Advanced Baseline Imager (ABI). These are used to classify CloudSat/CALIPSO observed multi-layer clouds. Due to the imbalanced nature of the data, we investigate the adoption of conventional resampling methods, as well as deep learning methods with data augmentation. In our experiments, we utilize the random forest classifier and Multilayer perceptron classifier with data augmentation methods to reduce the class imbalance during training. With these approaches, we achieve a classification accuracy of 83.6% without exploiting any ancillary information.

machine learning↗

LANDSAT image differencing as an automated land cover change detection technique

Image differencing was investigated as a technique for use with LANDSAT digital data to delineate areas of land cover change in an urban environment. LANDSAT data collected in April 1973 and April 1975 for Austin, Texas, were geometrically corrected and precisely registered to United States Geological Survey 7.5-minute quadrangle maps. At each pixel location reflectance values for the corresponding bands were subtracted to produce four difference images. Areas of major reflectance differences are isolated by thresholding each of the difference images. The resulting images are combined to obtain an image data set to total change. These areas of reflectance differences were found, in general, to correspond to areas of land cover change. Information on areas of land cover change was incorporated into a procedure to mask out all nonchange areas and perform an unsupervised classification only for data in the change areas. This procedure identified three broad categories: (1) areas of high reflectance (construction or extractive), (2) changes in agricultural areas, and (3) areas of confusion between agricultural and other areas.

Stauffer, M. L.↗

Future of Big Earth Data Analytics

The state of the art of Big Earth Data Analytics can be expected to evolve rapidly in the coming years. The forces driving evolution come from both growth in the data and advancement in the field of data analytics. In the data area, advances in sensor instrumentation and platform miniaturization are increasing both data resolution and coverage, resulting in enormous growth in data Volume. Increases in temporal resolution in particular also generate demands for higher data Velocity. At the same time, the proliferation of instruments and the platforms on which they reside is increasing the Variety of datasets. The Variety increase in turn leads to questions about the Veracity of the data. In the algorithm area, powerful machine learning methods are coming to the fore, particularly Deep Neural Networks. These are powerful at detecting interesting features in the data, integrating many different measurements (i.e., data fusion), and classification problems. However, they are still challenging when seeking explanations of how natural or socio-economic phenomena work using Earth Observations. Thus, classical analysis techniques will remain relevant when the emphasis is on forming or testing explanations, as well as to support interactive data exploration.

Lynnes, Christopher↗

Automatic classification of time-variable X-ray sources

To maximize the discovery potential of future synoptic surveys, especially in the field of transient science, it will be necessary to use automatic classification to identify some of the astronomical sources. The data mining technique of supervised classification is suitable for this problem. Here, we present a supervised learning method to automatically classify variable X-ray sources in the Second XMM-Newton Serendipitous Source Catalog (2XMMi-DR2). Random Forest is our classifier of choice since it is one of the most accurate learning algorithms available. Our training set consists of 873 variable sources and their features are derived from time series, spectra, and other multi-wavelength contextual information. The 10 fold cross validation accuracy of the training data is ∼97% on a 7 class data set. We applied the trained classification model to 411 unknown variable 2XMM sources to produce a probabilistically classified catalog. Using the classification margin and the Random Forest derived outlier measure, we identified 12 anomalous sources, of which 2XMM J180658.7–500250 appears to be the most unusual source in the sample. Its X-ray spectra is suggestive of a ultraluminous X-ray source but its variability makes it highly unusual. Machine-learned classification and anomaly detection will facilitate scientific discoveries in the era of all-sky surveys.

79 ASTRONOMY AND ASTROPHYSICS↗

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↗

Identification and interpretation of patterns in rocket engine data: Artificial intelligence and neural network approaches

This paper describes an expert system which is designed to perform automatic data analysis, identify anomalous events, and determine the characteristic features of these events. We have employed both artificial intelligence and neural net approaches in the design of this expert system. The artificial intelligence approach is useful because it provides (1) the use of human experts' knowledge of sensor behavior and faulty engine conditions in interpreting data; (2) the use of engine design knowledge and physical sensor locations in establishing relationships among the events of multiple sensors; (3) the use of stored analysis of past data of faulty engine conditions; and (4) the use of knowledge-based reasoning in distinguishing sensor failure from actual faults. The neural network approach appears promising because neural nets (1) can be trained on extremely noisy data and produce classifications which are more robust under noisy conditions than other classification techniques; (2) avoid the necessity of noise removal by digital filtering and therefore avoid the need to make assumptions about frequency bands or other signal characteristics of anomalous behavior; (3) can, in effect, generate their own feature detectors based on the characteristics of the sensor data used in training; and (4) are inherently parallel and therefore are potentially implementable in special-purpose parallel hardware.

Ali, Moonis↗

IDENTIFYING ROOF FALL PREDICTORS USING FUZZY CLASSIFICATION

Microseismic monitoring involves placing geophones on the rock surfaces of a mine to record seismic activity. Classification of microseismic mine data can be used to predict seismic events in a mine to mitigate mining hazards, such as roof falls, where properly bolting and bracing the roof is often an insufficient method of preventing weak roofs from destabilizing. In this study, six months of recorded acoustic waveforms from microseismic monitoring in a Pennsylvania limestone mine were analyzed using classification techniques to predict roof falls. Fuzzy classification using features selected for computational ease was applied on the mine data. Both large roof fall events could be predicted using a Roof Fall Index (RFI) metric calculated from the results of the fuzzy classification. RFI was successfully used to resolve the two significant roof fall events and predicted both events by at least 15 hours before visual signs of the roof falls were evident.

42 ENGINEERING↗

Land use in the Paraiba Valley through remotely sensed data

A methodology for land use survey was developed and land use modification rates were determined using LANDSAT imagery of the Paraiba Valley (state of Sao Paulo). Both visual and automatic interpretation methods were employed to analyze seven land use classes: urban area, industrial area, bare soil, cultivated area, pastureland, reforestation and natural vegetation. By means of visual interpretation, little spectral differences are observed among those classes. The automatic classification of LANDSAT MSS data using maximum likelihood algorithm shows a 39% average error of omission and a 3.4% error of inclusion for the seven classes. The complexity of land uses in the study area, the large spectral variations of analyzed classes, and the low resolution of LANDSAT data influenced the classification results.

Dejesusparada, N.↗

Assessment of satellite and aircraft multispectral scanner data for strip-mine monitoring

The application of LANDSAT multispectral scanner data to describe the mining and reclamation changes of a hilltop surface coal mine in the rugged, mountainous area of eastern Kentucky is presented. Original single band satellite imagery, computer enhanced single band imagery, and computer classified imagery are presented for four different data sets in order to demonstrate the land cover changes that can be detected. Data obtained with an 11 band multispectral scanner on board a C-47 aircraft at an altitude of 3000 meters are also presented. Comparing the satellite data with color, infrared aerial photography, and ground survey data shows that significant changes in the disrupted area can be detected from LANDSAT band 5 satellite imagery for mines with more than 100 acres of disturbed area. However, band-ratio (bands 5/6) imagery provides greater contrast than single band imagery and can provide a qualitative level 1 classification of the land cover that may be useful for monitoring either the disturbed mining area or the revegetation progress. However, if a quantitative, accurate classification of the barren or revegetated classes is required, it is necessary to perform a detailed, four band computer classification of the data.

Spisz, E. W.↗

Feature Extraction: Improving Remote Sensor Classification of Non-Proliferation

This research focuses on developing algorithms for nuclear non-proliferation detection using remote sensor modeling. To improve the performance of classification models, we implemented a data pipeline with feature extraction. This pipeline takes raw data and transforms it into smaller data points called features that still describe the model. Improving this classification works towards the departments of energy’s missions of ensuring American’s security and prosperity by creating technology that addresses nuclear challenges. To conduct this analysis, we used the Python programming language and some key packages, including tsfresh and TSFEL. Originally tsfresh was selected because it has the most statistical features out of all the packages. Later TSFEL was incorporated due to the additional features it can extract from data, such as temporal and spectral. However, feature extraction becomes challenging in the presence of missing values. In this case, two additional Python packages were added to our workflow, NumPy and pandas, allowing for the feature extraction process to handle unknown values. Our data pipeline was tested on data collected from a simulation that describes the process state of a physical example. The results show the pipeline’s capability to consume and extract a total 17 features from tabular data. Future work includes producing classifications using decision tree-based models such as XGBoost and improving data collection by analyzing feature importance.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Fanning - A classification algorithm for mixture landscapes applied to Landsat data of Maine forests

It is pointed out that typical landscapes include a relatively small number of 'pure' land cover types which combine in various proportions to form a myriad of mixture types. Most Landsat classifications algorithms used today require a separate user specification for each category, including mixture categories. Attention is given to a simpler approach, which would require the user to specify only the 'pure' types. Mixture pixels would be classified on the basis of the proportion of the area covered by each pure type within the pixel. The 'fanning' algorithm quantifies varying proportions of two 'pure' land cover types in selected mixture pixels. This algorithm was applied to 200,000 ha of forest land in Maine, taking into account a comparison with standard inventory information. Results compared well with a discrete categories classification of the same area.

Ungar, S. G.↗

Satellite-based snow identification and its impact on monitoring photovoltaic systems

Earth observation allows the separation of snow cover and cloudiness using multispectral measurements. Several satellite-based snow monitoring services are available, ranging from regional to world-wide scales. Using these data enables photovoltaic (PV) plant management to differentiate between failures due to snow coverage on a PV system and other error sources. Additionally, yield estimates for solar siting are improved. This paper presents a validation study from January to April 2006 comparing satellite-based datasets with ground measurements from German and Swiss meteorological stations. A false alarm rate, an error due to irradiance underestimation, the availability of daily data, and the classification accuracy are introduced as quality metrics. Compared to Switzerland, generally a higher accuracy is found in all datasets for Southern Germany. The most significant difference among the datasets is found in the error pattern shifting from too much snow (which results in an error due to underestimation of irradiance) to too little snow detection, causing a false alarm in PV monitoring. Overall, the data records of the Land Surface Analysis Satellite Application Facility (LSA SAF), the German Aerospace Center (DLR) and the Interactive Multisensor Snow and Ice Mapping System (IMS) are found to be most suitable for solar energy purposes. The IMS dataset has a low false alarm rate (4%) and a good data availability (100%) making it a good choice for power plant monitoring, but the error due to underestimation relevant in site auditing is large with 59%. If a cumulative snow cover algorithm is applied to achieve information every day as needed both for power plant monitoring and site auditing, both the DLR and the LSA SAF datasets are comparable with classification accuracies of 70%, false alarm rates of 37% and 34%, respectively, and errors due to irradiance underestimation in 26% and 27% of all coincidences. (author)

14 SOLAR ENERGY↗

Molecular subtyping reveals immune alterations in IDH wild‐type lower‐grade diffuse glioma

Abstract Isocitrate dehydrogenase ( IDH ) wild‐type diffuse lower‐grade glioma (LGG) is usually associated with poor outcome, but there have been disputes over its clinical outcome and classification. We present here a robust gene expression‐based molecular classification of IDH wild‐type diffuse LGG into two subtypes with distinct biological and clinical features. A discovery cohort of 49 IDH wild‐type diffuse LGGs from the Chinese Glioma Genome Atlas (CGGA) was subjected to clustering and function analysis. Seventy‐three tumors from The Cancer Genome Atlas (TCGA) were used to validate our findings. Consensus clustering of transcriptional data uncovered concordant classification of two robust and prognostically significant subtypes of IDH wild‐type LGG. Subtype 1, associated with poorer outcomes, was characterized by significantly higher immune and cytolytic scores, M2 macrophages, and up‐regulation of immune exhaustion markers, while Subtype 2, which had elevated lymphocytes and plasma cells, showed relatively favorable survival. Somatic alteration analysis revealed that Subtype 1 showed more frequently deleted regions, such as the locus of CDKN2A / CDKN2B , DMRTA1 , C9orf53 , and MTAP . Furthermore, we developed and validated a five‐gene signature for better application of this acquired stratification. Our data demonstrate the biological and prognostic heterogeneity within IDH wild‐type diffuse LGGs and deepen our molecular understandi‐g of this tumor entity. © 2020 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd.

Wu, Fan↗

Classification of earth terrain using polarimetric synthetic aperture radar images

Supervised and unsupervised classification techniques are developed and used to classify the earth terrain components from SAR polarimetric images of San Francisco Bay and Traverse City, Michigan. The supervised techniques include the Bayes classifiers, normalized polarimetric classification, and simple feature classification using discriminates such as the absolute and normalized magnitude response of individual receiver channel returns and the phase difference between receiver channels. An algorithm is developed as an unsupervised technique which classifies terrain elements based on the relationship between the orientation angle and the handedness of the transmitting and receiving polariation states. It is found that supervised classification produces the best results when accurate classifier training data are used, while unsupervised classification may be applied when training data are not available.

Lim, H. H.↗

Comparative utility of LANDSAT-1 and Skylab data for coastal wetland mapping and ecological studies

Skylab 190-A photography and LANDSAT-1 analog data have been analyzed to determine coastal wetland mapping potential as a near term substitute for aircraft data and as a long term monitoring tool. The level of detail and accuracy of each was compared. Skylab data provides more accurate classification of wetland types, better delineation of freshwater marshes and more detailed analysis of drainage patterns. LANDSAT-1 analog data is useful for general classification, boundary definition and monitoring of human impact in wetlands.

Anderson, R.↗

Classification improvement by optimal dimensionality reduction when training sets are of small size

A computer simulation was performed to test the conjecture that, when the sizes of the training sets are small, classification in a subspace of the original data space may give rise to a smaller probability of error than the classification in the data space itself; this is because the gain in the accuracy of estimation of the likelihood functions used in classification in the lower dimensional space (subspace) offsets the loss of information associated with dimensionality reduction (feature extraction). A number of pseudo-random training and data vectors were generated from two four-dimensional Gaussian classes. A special algorithm was used to create an optimal one-dimensional feature space on which to project the data. When the sizes of the training sets are small, classification of the data in the optimal one-dimensional space is found to yield lower error rates than the one in the original four-dimensional space.

Starks, S. A.↗