SEARCH · Search NASA
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.
On-line handwritten alphanumeric character recognition system and learning algorithm for multiclass pattern classification
The Atmospheric Radiation Measurement (ARM) Data Center is responsible for the timely collection, archival, and curation of science data products. These products are freely available through an online data repository. Metadata creation is paramount for scientific users to find and access over seven petabytes of atmospheric science data. The hierarchical metadata structure allows users to search for information at both broad and narrow levels. This project aims to leverage 30 years’ worth of manually created metadata to enable machine predictions of broad-term classifications from narrow-term descriptions. These classification predictions would assist metadata coordinators with their term selections. This paper discusses the cleaning and preprocessing of the training data, the pipeline developed to determine the best model for this task, and the creation of an API metadata classifier for ARM measurement metadata. Our results show that the Linear Support Vector Classification (LinearSVC) algorithm, along with the Term Frequency – Inverse Document Frequency (TF-IDF) vectorizer, is well-suited for our multi-class classification task. Lengthier input training data led to better results, and artificial balancing was unnecessary for this particular use case. This predictive classifier enhances efficiency in metadata creation, as well as supports greater consistency and accuracy in metadata tagging.
Where spatial boundaries between phenomena are diffuse, classification methods which construct mutually exclusive clusters seem inappropriate. The Fuzzy c-means (FCM) algorithm assigns each observation to all clusters, with membership values as a function of distance to the cluster center. The FCM algorithm is applied to AVHRR data for the purpose of classifying polar clouds and surfaces. Careful analysis of the fuzzy sets can provide information on which spectral channels are best suited to the classification of particular features, and can help determine likely areas of misclassification. General agreement in the resulting classes and cloud fraction was found between the FCM algorithm, a manual classification, and an unsupervised maximum likelihood classifier.
Where spatial boundaries between phenomena are diffuse, classification methods which construct mutually exclusive clusters seem inappropriate. The Fuzzy c-means (FCM) algorithm assigns each observation to all clusters, with membership values as a function of distance to the cluster center. The FCM algorithm is applied to AVHRR data for the purpose of classifying polar clouds and surfaces. Careful analysis of the fuzzy sets can provide information on which spectral channels are best suited to the classification of particular features, and can help determine like areas of misclassification. General agreement in the resulting classes and cloud fraction was found between the FCM algorithm, a manual classification, and an unsupervised maximum likelihood classifier.
Logging operations are widespread across the Oregon Coast Range and conventional logging practices pose a risk of contamination to surface water quality. The NASA DEVELOP Oregon Coast Ecological Conservation team partnered with nonprofit Oregon Wild to quantify the extent of clearcutting and commercial thinning in 80 Coast Range drinking watersheds between 2000 and 2022. This project used all available Landsat data from 1997 through June 2023 in Google Earth Engine. Sensors used include Landsat 5 Thematic Mapper, Landsat 7 Enhanced Thematic Mapper Plus, Landsat 8 Operational Land Imager, and Landsat 9 Operational Land Imager-2. The Continuous Change Detection and Classification (CCDC) algorithm was used with Landsat observations to identify clearcutting patches. Percent change in summer median Landsat Normalized Difference Vegetation Index (NDVI) images were used to identify areas of forest disturbance including commercial thinning. The team concluded that logging, including both clearcutting and commercial thinning, impacted 31% of forested area in drinking watersheds and the intensity of logging remained consistent from year to year. Clearcutting occurred primarily on private land while commercial thinning occurred primarily on state and federal lands. This study showed that CCDC effectively identifies clearcutting, and percent change in NDVI successfully identifies disturbances including commercial thinning. Key constraints included the lack of field validation data and the inability to attribute disturbances to logging with certainty. Ultimately, this study identified the drinking watersheds and communities most likely to be impacted by logging activity. These results can inform legislation aimed at balancing the commercial and environmental benefits of forestlands.
Purpose Integrating artificial intelligence in cancer diagnostics has improved tumor classification beyond rule-based systems. Despite these advancements, these models may still encode demographic biases. We conducted a large-scale, applied bias-probing study of a deep learning–based cancer site classifier to quantify race information encoded in document embeddings. We then evaluated how performance changes when race-correlated embedding dimensions are removed in a post-training sensitivity analysis. Methods The cancer site classifier was trained using 3.5 million electronic cancer pathology reports from six of the National Cancer Institute's SEER registries. We trained a hierarchical self-attention network to generate 400-dimensional document embeddings. These embeddings were used to train two downstream, gradient-boosted decision tree classifiers: one to classify the cancer sites and another to predict racial categories. We identified overlapping features by intersecting the top 50 feature-importance rankings from the site and race models and computed their cumulative feature importance in each model. As a post hoc sensitivity analysis, we progressively pruned these overlapping dimensions, retrained the site model, and compared overall macro-F1 and accuracy, race-stratified macro-F1, and group fairness metrics on the basis of demographic parity and equalized odds before and after pruning. Results The analysis revealed minimal feature overlap between the cancer site and race prediction models, and the cumulative importance scores indicated a negligible influence of racial information on clinical predictions. Post-training pruning of overlapping features did not compromise the models' diagnostic accuracy, with a 0.07% loss in accuracy. Conclusion Our findings demonstrate that HiSAN-generated embeddings from SEER data can be used effectively in cancer site classification without significant demographic bias influencing the outcomes. Post-training pruning therefore functions as a practical audit and sensitivity check.
Equivalence relations for optimal control problems classification and algorithm construction
Two simple algorithms for classification of sedimentation levels in water and for delineation of active strip mines are in use as part of the development of a more general resource management information system. ERTS MSS CCT's are processed so that each pixel in each channel is geographically referenced and can be accessed individually during whole frame, multi-channel analysis or partial frame analysis. The sedimentation analysis clearly separates classes representing the turbid Anacostia water, the less disturbed Potomac (really), and mud flats resulting from effluent of a major sewage treatment plant. Mud flats of organic or mineral origin are easily distinguished.
Explore the source record for details and available documents.
The data collections archived and distributed by the GES DISC NASA data center are widely utilized for various Earth Science studies. As these collections are created, many research works are published regarding these collections' algorithms, their validation, and their applications. As NASA data centers collect these publications for public use, it is helpful to categorize them based on how they relate to their associated datasets. Specifically, whether the publication linked to the GES DISC dataset is using it for applicational research, describing the algorithm used for the dataset creation, validating the dataset, or providing a general overview of the data collection. Currently, this process requires simple manual labeling, and as such, it may be possible to solve via automation. To approach this problem, machine learning classifiers were developed to predict a publication's category. Manually labeled publications were used as the training data for the supervised machine learning algorithms, specifically Random Forest and Multinomial Naïve Bayes. After balancing the dataset and implementing the Multinomial Naïve Bayes algorithm, the classification accuracy achieved was substantially higher than the baseline accuracy, thus significantly improving the efficiency of publication labeling.
Mexico is one of the countries with great potential for the UN's Reducing Emissions from Deforestation and Forest Degradation (REDD+) program, a key nature-based solution for the forest sector. To monitor carbon stock changes, there is a growing demand for unbiased Monitoring Reporting Verification (MRV) systems to facilitate effective forest management and climate change mitigation strategies. Remote sensing-based national aboveground biomass density (AGBD) estimation over Mexico is scarce and often limited to one-time static mapping, leading to spatiotemporal inconsistency in inputs. As an effort under NASA's Carbon Monitoring System (CMS) program, we have developed a remote sensing-based approach to create consistent historical AGBD maps of Mexico using multi-stream remote sensing data, including spaceborne lidar GEDI and long-term Landsat time series, as well as topographic information. We employ the continuous change detection and classification (CCDC) algorithm for temporal modeling of Landsat surface reflectance, followed by the inference of forest AGBD using a random forest machine learning algorithm with the temporal information of land surface dynamics extracted by the CCDC as input. GEDI provides unprecedented forest structure and AGBD sampling datasets for model training and validation practices. In this presentation, we share the progress made in developing a spatially explicit mapping of historical AGBD changes associated with land surface changes and post-disturbance landscapes.
A lossless image-data-compression algorithm intended specifically for application to classification-map data is based on prediction, context modeling, and entropy coding. The algorithm was formulated, in consideration of the differences between classification maps and ordinary images of natural scenes, so as to be capable of compressing classification- map data more effectively than do general-purpose image-data-compression algorithms. Classification maps are typically generated from remote-sensing images acquired by instruments aboard aircraft (see figure) and spacecraft. A classification map is a synthetic image that summarizes information derived from one or more original remote-sensing image(s) of a scene. The value assigned to each pixel in such a map is the index of a class that represents some type of content deduced from the original image data for example, a type of vegetation, a mineral, or a body of water at the corresponding location in the scene. When classification maps are generated onboard the aircraft or spacecraft, it is desirable to compress the classification-map data in order to reduce the volume of data that must be transmitted to a ground station.
The ocean has been traditionally viewed as a 2 class system. Morel and Prieur (1977) classified ocean water according to the dominant absorbent particle suspended in the water column. Case 1 is described as having a high concentration of phytoplankton (and detritus) relative to other particles. Conversely, case 2 is described as having inorganic particles such as suspended sediments in high concentrations. Little work has gone into the problem of mixing bio-optical models for these different water types. An approach is put forth here to blend bio-optical algorithms based on a fuzzy classification scheme. This scheme involves two procedures. First, a clustering procedure identifies classes and builds class statistics from in-situ optical measurements. Next, a classification procedure assigns satellite pixels partial memberships to these classes based on their ocean color reflectance signature. These membership assignments can be used as the basis for a weighting retrievals from class-specific bio-optical algorithms. This technique is demonstrated with in-situ optical measurements and an image from the SeaWiFS ocean color satellite.
The Feature Identification and Location Experiment (FILE) flown on the Shuttle STS-2 mission November 12-14, 1981, tested a technique for autonomous real-time classification of selected earth features, i.e., water; bare land; vegetation; and clouds, snow, and ice. A second instrument, designed for aircraft flights, flew over regions of the west and east coasts of the United States and across the country. In each instrument, two bore-sighted CCD cameras image earth scenes in two spectral bands. Each camera includes a 100-element by 100-element detector array, and classification circuits. A simple algorithm and logic circuit provides classification decisions within a few microseconds. The experiment records the number of picture elements (pixels) representing each feature and the reflected solar radiation for each band. After flight, pixel-by-pixel classification images are constructed and compared with 70-mm color photographs taken simultaneously with the CCD-camera data.
Broadband U, V, and x photometry together with IRAS asteroid albedos have been used to construct an asteroid classification system. The system is based on three parameters (U-V and v-x color indices and visual geometric albedo), and it is able to place 96 percent of the present sample of 357 asteroids into 11 taxonomic classes. It is noted that all but one of these classes are analogous to those previously found using other classification schemes. The algorithm is shown to account for the observational uncertainties in each of the classification parameters.
This paper presents a signal processing system that: (1) provides an efficient and reliable instrument for investigating the activity of neuronal assemblies in the brain; and (2) demonstrates the feasibility of generating the command signals of prostheses using the activity of relevant neurons in disabled subjects. The system operates online, in a fully automated manner and can recognize the transient waveforms of several neurons in extracellular neurophysiological recordings. Optimal algorithms for detection, classification, and resolution of overlapping waveforms are developed and evaluated. Full automation is made possible by an algorithm that can set appropriate decision thresholds and an algorithm that can generate templates on-line. The system is implemented with a fast IBM PC compatible processor board that allows on-line operation.
A potential hardware organization for the Flexible Processor Array is presented. An algorithm that implements a contextual classifier for remote sensing data analysis is given, along with uniprocessor classification algorithms. The Flexible Processor algorithm is provided, as are simulated timings for contextual classifiers run on the Flexible Processor Array and another system. The timings are analyzed for context neighborhoods of sizes three and nine.
We describe an application of a scale-space clustering algorithm to the classification of a multispectral and polarimetric SAR image of an agricultural site. After the initial polarimetric and radiometric calibration and noise cancellation, we extracted a 12-dimensional feature vector for each pixel from the scattering matrix. The clustering algorithm was able to partition a set of unlabeled feature vectors from 13 selected sites, each site corresponding to a distinct crop, into 13 clusters without any supervision. The cluster parameters were then used to classify the whole image. The classification map is much less noisy and more accurate than those obtained by hierarchical rules. Starting with every point as a cluster, the algorithm works by melting the system to produce a tree of clusters in the scale space. It can cluster data in any multidimensional space and is insensitive to variability in cluster densities, sizes and ellipsoidal shapes. This algorithm, more powerful than existing ones, may be useful for remote sensing for land use.