Finding Faults in PV Systems: Supervised and Unsupervised Dictionary Learning With SSTDR
Not provided.
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.
Not provided.
ABSTRACT The Amazon rainforest has been subjected to high rates of deforestation, mostly for pasturelands, over the last few decades. This change in plant cover is known to alter the soil microbiome and the functions it mediates, but the genomic changes underlying this response are still unresolved. In this study, we used a combination of deep shotgun metagenomics complemented by a supervised machine learning approach to compare the metabolic strategies of tropical soil microbial communities in pristine forests and long‐term established pastures in the Amazon. Machine learning‐derived metagenome analysis indicated that microbial community structures (bacteria, archaea and viruses) and the composition of protein‐coding genes were distinct in each plant cover type environment. Forest and pasture soils had different genomic diversities for the above three taxonomic groups, characterised by their protein‐coding genes. These differences in metagenome profiles in soils under forests and pastures suggest that metabolic strategies related to carbohydrate and energy metabolisms were altered at community level. Changes were also consistent with known modifications to the C and N cycles caused by long‐term shifts in aboveground vegetation and were also associated with several soil physicochemical properties known to change with land use, such as the C/N ratio, soil temperature and exchangeable acidity. In addition, our analysis reveals that these alterations in land use can also result in changes to the composition and diversity of the soil DNA virome. Collectively, our study indicates that soil microbial communities shift their overall metabolic strategies, driven by genomic alterations observed in pristine forests and long‐term established pastures with implications for the C and N cycles.
We describe how to use several machine learning techniques organized in a learning pipeline to segment and identify cell membrane structures from cryo electron tomograms. These tomograms are difficult to analyze with traditional segmentation tools. The learning pipeline in our approach starts from supervised learning via a special convolutional neural network trained with simulated data. It continues with semi-supervised reinforcement learning and/or a region merging technique that tries to piece together disconnected components belonging to the same membrane structure. A parametric or non-parametric fitting procedure is then used to enhance the segmentation results and quantify uncertainties in the fitting. Domain knowledge is used in generating the training data for the neural network and in guiding the fitting procedure through the use of appropriately chosen priors and constraints. We demonstrate that the approach proposed here works well for extracting membrane surfaces in two real tomogram datasets.
We report faults in Heating, Ventilation, and Air Conditioning (HVAC) systems of buildings result in significant energy waste in building operation. With fast-growing sensing data availability and advancement in computing, computational modeling has demonstrated strong capability to detect and diagnose HVAC system faults, hence, ensuring efficient building operation. This paper comprehensively reviews the state-of-the-art computing-based fault detection and diagnosis (FDD) for HVAC systems. Overall, the reviewed computing-based FDD methods are classified as two major approaches: knowledge-based and data-driven approaches. We then identify multiple important topics, including data availability, training data size, data quality, approach generality, capability, interpretability, and required modeling efforts, along with corresponding metrics to summarize the most updated FDD development. Generally, the knowledge-based approaches are further divided as physics-based modeling, Diagnostic Bayesian Network, and performance indicator-based methods while data-driven approaches include supervised learning, unsupervised learning, and regression and statistics-based methods. State-of-the-art FDD development, remaining challenges, and future research directions are further discussed to push forward FDD in practice. Availability of fault data, capability of existing methods to deal with complex fault situations (such as simultaneous faults), modeling interpretability for data-driven methods, and required engineering efforts for physics-based methods are identified as remaining challenges in FDD development. Improving modeling fidelity and reducing modeling efforts are essential for applying physics-based methods in real buildings. Meanwhile, addressing fault data availability, increasing algorithm adaptability, and handling multiple faults are essential to further enhance the applicability of data-driven FDD approaches.
An extensive experimental and theoretical study of the Kα and Kβ high-resolution X-ray emission spectroscopy (XES) of sulfur-bearing systems is presented here. This study encompasses a wide range of organic and inorganic compounds, including numerous experimental spectra from both prior published work and new measurements. Employing a linear-response time-dependent density functional theory (LR-TDDFT) approach, strong quantitative agreement is found in the calculation of energy shifts of the core-to-core Kα as well as the full range of spectral features in the valence-to-core Kβ spectrum. The ability to accurately calculate the sulfur Kα energy shift supports the use of sulfur Kα XES as a bulk-sensitive tool for assessing sulfur speciation. The fine structure of the sulfur Kβ spectrum, in conjunction with the theoretical results, is shown to be sensitive to the local electronic structure including effects of symmetry, ligand type and number, and, in the case of organosulfur compounds, to the nature of the bonded organic moiety. This agreement between theory and experiment, augmented by the potential for high-access XES measurements with the latest generation of laboratory-based spectrometers, demonstrates the possibility of broad analytical use of XES for sulfur and nearby third-row elements. The effective solution of the forward problem, i.e., successful prediction of detailed spectra from known molecular structure, also suggests future use of supervised machine learning approaches to experimental inference, as has seen recent interest for interpretation of X-ray absorption near-edge structure (XANES).
In this study, multiple instance learning (MIL) algorithms to automatically perform root detection and segmentation in minirhizotron imagery using only image-level labels are proposed. Root and soil characteristics vary from location to location, and thus, supervised machine learning approaches that are trained with local data provide the best ability to identify and segment roots in minirhizotron imagery. However, labeling roots for training data (or otherwise) is an extremely tedious and time-consuming task. This paper aims to address this problem by labeling data at the image level (rather than the individual root or root pixel level) and train algorithms to perform individual root pixel level segmentation using MIL strategies. Three MIL methods (multiple instance adaptive cosine coherence estimator, multiple instance support vector machine, multiple instance learning with randomized trees) were applied to root detection and compared to non-MIL approaches. The results show that MIL methods improve root segmentation in challenging minirhizotron imagery and reduce the labeling burden. In our results, multiple instance support vector machine outperformed other methods. The multiple instance adaptive cosine coherence estimator algorithm was a close second with an added advantage that it learned an interpretable root signature which identified the traits used to distinguish roots from soil and did not require parameter selection.
Despite several attempts, automated detection of microaneurysm (MA) from digital fundus images still remains to be an open issue. This is due to the subtle nature of MAs against the surrounding tissues. In this paper, the microaneurysm detection problem is modeled as finding interest regions or blobs from an image and an automatic local-scale selection technique is presented. Several scale-adapted region descriptors are then introduced to characterize these blob regions. A semi-supervised based learning approach, which requires few manually annotated learning examples, is also proposed to train a classifier to detect true MAs. The developed system is built using only few manually labeled and a large number of unlabeled retinal color fundus images. The performance of the overall system is evaluated on Retinopathy Online Challenge (ROC) competition database. A competition performance measure (CPM) of 0.364 shows the competitiveness of the proposed system against state-of-the art techniques as well as the applicability of the proposed features to analyze fundus images.
Abstract Invasive plant pathogenic fungi have a global impact, with devastating economic and environmental effects on crops and forests. Biosurveillance, a critical component of threat mitigation, requires risk prediction based on fungal lifestyles and traits. Recent studies have revealed distinct genomic patterns associated with specific groups of plant pathogenic fungi. We sought to establish whether these phytopathogenic genomic patterns hold across diverse taxonomic and ecological groups from the Ascomycota and Basidiomycota, and furthermore, if those patterns can be used in a predictive capacity for biosurveillance. Using a supervised machine learning approach that integrates phylogenetic and genomic data, we analyzed 387 fungal genomes to test a proof-of-concept for the use of genomic signatures in predicting fungal phytopathogenic lifestyles and traits during biosurveillance activities. Our machine learning feature sets were derived from genome annotation data of carbohydrate-active enzymes (CAZymes), peptidases, secondary metabolite clusters (SMCs), transporters, and transcription factors. We found that machine learning could successfully predict fungal lifestyles and traits across taxonomic groups, with the best predictive performance coming from feature sets comprising CAZyme, peptidase, and SMC data. While phylogeny was an important component in most predictions, the inclusion of genomic data improved prediction performance for every lifestyle and trait tested. Plant pathogenicity was one of the best-predicted traits, showing the promise of predictive genomics for biosurveillance applications. Furthermore, our machine learning approach revealed expansions in the number of genes from specific CAZyme and peptidase families in the genomes of plant pathogens compared to non-phytopathogenic genomes (saprotrophs, endo- and ectomycorrhizal fungi). Such genomic feature profiles give insight into the evolution of fungal phytopathogenicity and could be useful to predict the risks of unknown fungi in future biosurveillance activities.
Using a semi-supervised machine learning approach we present a real-time anomaly detection system based on an autoencoder used for online data quality monitoring of the CMS electromagnetic calorimeter operating at the CERN LHC. We introduce a novel method that maximizes the anomaly detection performance making use of the time-dependence of anomalies and the spatial variations in the detector response. The autoencoder-based system efficiently detects anomalies in real time and maintains a very low false discovery rate. We validate the performance of this novel system with anomalies from LHC collision data taken in 2018 and 2022. In addition, results are presented after deploying the autoencoder-based system in the CMS online Data Quality Monitoring workflow at the beginning of LHC Run 3 resulting in the system to detect issues that were missed by the existing system.
Understanding the origins of enhanced reactivity of supported, subnanometer in size, metal oxide clusters is challenging due to the scarcity of methods capable to extract atomic-level information from the experimental data. Due to both the sensitivity of X-ray absorption near edge structure (XANES) spectroscopy to the local geometry around metal ions and reliability of theoretical spectroscopy codes for modeling XANES spectra, supervised machine learning approach has become a powerful tool for extracting structural information from the experimental spectra. Here, we present the application of this method to grazing incidence XANES spectra of size-selective Cu oxide clusters on flat support, measured in operando conditions of the methanation reaction. We demonstrate that the convolution neural network can be trained on theoretical spectra and utilized to “invert” experimental XANES data to obtain structural descriptors—the Cu–Cu coordination numbers. As a result, we were able to distinguish between different structural motifs (Cu 2 O-like and CuO-like) of Cu oxide clusters, transforming in reaction conditions, and reliably evaluate average cluster sizes, with important implications for the understanding of structure, composition, and function relationships in catalysis.
Non-negative Matrix Factorization (NMF) has proven to be a powerful unsupervised learning method for uncovering hidden features in complex and noisy datasets with applications in data mining, text recognition, dimension reduction, face recognition, anomaly detection, blind source separation, and many other fields. An important input for NMF is the latent dimensionality of the data, that is, the number of hidden features, K, present in the explored dataset. Unfortunately, and this quantity is rarely known a priori. The existing methods for determining latent dimensionality, such as Automatic Relevance Determination (ARD), are mostly heuristic and utilize different characteristics to estimate the number of hidden features. However, all of them require human presence to make a final determination of K. Here we utilize a supervised machine learning approach in combination with a recent method for model determination, called NMFk, to determine the number of hidden features automatically. NMFk performs a set of NMF simulations on an ensemble of matrices, obtained by bootstrapping the initial dataset, and estimates which K produces stable groups of latent features that reconstruct the initial dataset well. We then train a Multi-Layter Perceptron (MLP) classifier network to determine the correct number of latent features utilizing the statistics and characteristics of the NMF solution, obtained from NMFk. In order to train the MLP classifier, a training set of 58,660 matrices with predetermined latent features were factorized with NMFk. The MLP classifier in conjunction with NMFk maintains a greater than 95% success rate when applied to a held out test set. Additionally, when applied to two well-known benchmark datasets, the swimmer and MIT face data, NMFk/MLP correctly recovers the established number of hidden features. Finally, we compare the accuracy of our method to the ARD, AIC and Stability-based methods.
Galaxy-scale strong gravitational lensing can not only provide a valuable probe of the dark matter distribution of massive galaxies, but also provide valuable cosmological constraints, either by studying the population of strong lenses or by measuring time delays in lensed quasars. Due to the rarity of galaxy-scale strongly lensed systems, fast and reliable automated lens finding methods will be essential in the era of large surveys such as Large Synoptic Survey Telescope, Euclid and Wide-Field Infrared Survey Telescope. To tackle this challenge, we introduce CMU DeepLens, a new fully automated galaxy-galaxy lens finding method based on deep learning. This supervised machine learning approach does not require any tuning after the training step which only requires realistic image simulations of strongly lensed systems. We train and validate our model on a set of 20 000 LSST-like mock observations including a range of lensed systems of various sizes and signal-to-noise ratios (S/N). We find on our simulated data set that for a rejection rate of non-lenses of 99 per cent, a completeness of 90 per cent can be achieved for lenses with Einstein radii larger than 1.4 arcsec and S/N larger than 20 on individual g-band LSST exposures. Finally, we emphasize the importance of realistically complex simulations for training such machine learning methods by demonstrating that the performance of models of significantly different complexities cannot be distinguished on simpler simulations. We make our code publicly available at https://github.com/McWilliamsCenter/CMUDeepLens.
This tool takes a genome assembly as input and predicts per-site nucleosome occupancy as output. Trained on physical maps of nucleosome binding preferences across the fungal kingdom, NuclPred can be applied broadly across fungi (and other eukaryotes). This breadth, combined with its accuracy, means it could have both basic and applied biological implications, for example in understanding eukaryotic gene regulation and genetic engineering. Almost universally across eukaryotes, nucleosomes - each wrapping ~150 base pairs of DNA - serve to package DNA inside the nucleus, with major consequences on DNA access, gene activity and DNA integration. NuclPred was generated using a supervised deep learning approach combining convolutional and recurrent neural networks to take DNA features (nucleotides, GC content and structural information) as input, then use that information to predict the physical attractiveness DNA sequences might have for forming nucleosomes. With this information at hand, researchers can design more efficient CRISPR constructs, explore the interplay between DNA signatures and other regulators impact nucleosome locations, predict expression patterns, etc. This tool will be published as part of a manuscript currently under revision at iScience (draft attached).
In the atmosphere, secondary organic aerosols (SOA) are often the major components of fine particulate matter and interact with clouds and radiation. SOA comprises a mixture of thousands of organic compounds. There is tremendous complexity and uncertainty in understanding SOA formation, since it is formed by oxidation and gas to particle conversion of a variety of sources: natural biogenic, anthropogenic (vehicles, cooking coal combustion) and biomass burning. The Aerosol Mass Spectrometer (AMS) produces multidimensional chemical information about SOA but analyzing this data to understand SOA sources relies on time consuming analyses (~months to years) such as the positive matrix factorization (PMF). PMF also becomes difficult for aircraft data where signal to noise ratio is weaker. There is a critical need to develop fast machine learning techniques that can analytically provide information about SOA sources using AMS data on the same timescales as the data is being collected (~minutes). We apply a machine learning supervised classification approach: the multinomial logistic regression to rapidly classify AMS data obtained from aircraft measurements.
This project developed a novel statistical understanding of compression analytics (CA), which has challenged and clarified some core assumptions about CA, and enabled the development of novel techniques that address vital challenges of national security. Specifically, this project has yielded the development of novel capabilities including 1. Principled metrics for model selection in CA, 2. Techniques for deriving/applying optimal classification rules and decision theory to supervised CA, including how to properly handle class imbalance and differing costs of misclassification, 3. Two techniques for handling nonlocal information in CA, 4. A novel technique for unsupervised CA that is agnostic with regard to the underlying compression algorithm, 5. A framework for semisupervised CA when a small number of labels are known in an otherwise large unlabeled dataset. 6. The academic alliance component of this project has focused on the development of a novel exemplar-based Bayesian technique for estimating variable length Markov models (closely related to PPM [prediction by partial matching] compression techniques). We have developed examples illustrating the application of our work to text, video, genetic sequences, and unstructured cybersecurity log files.
Battery energy storage systems can be used for peak demand reduction in power systems, leading to significant economic benefits. Two practical challenges are 1) accurately determining the peak load days and hours and 2) quantifying and reducing uncertainties associated with the forecast in probabilistic risk measures for dispatch decision-making. In this study, we develop a supervised machine learning approach to generate 1) the probability of the next operation day containing the peak hour of the month and 2) the probability of an hour to be the peak hour of the day. Guidance is provided on preparation and augmentation of data as well as selection of machine learning models and decision-making thresholds. The proposed approach is applied to the Duke Energy Progress system and successfully captures 69 peak days out of 72 testing months with a 3% exceedance probability threshold. On 90% of the peak days, the actual peak hour is among the 2 h with the highest probabilities.
Dark matter halos are typically defined as spheres that enclose some overdensity, but these sharp, somewhat arbitrary boundaries introduce nonphysical artifacts such as backsplash halos, pseudo-volution, and an incomplete accounting of halo mass. A more physically motivated alternative is to define halos as the collection of particles that are physically orbiting within their potential well. However, existing methods to classify particles as orbiting or infalling suffer from trade-offs between accuracy, computational cost, and generalizability across cosmologies. We present an efficient, yet accurate, supervised machine learning approach using decision trees. The classification is based on only the particle radii and velocities at two epochs. Compared to detailed analysis of particle trajectories, we find that our model matches the classification of 97% of particles. Consequently, we are able to quickly and accurately reproduce the density profiles of the orbiting and infalling components out to many virial radii. We demonstrate that our model generalizes to a significantly different cosmology that lies outside the training data set. We make publicly available both our final model and the code to train similar models.
This study implemented a data-intensive supervised machine learning approach, through a quasi-experimental framework, with the objective of quantifying the impact of oil and gas operator-specific proprietary data on ML-based predictive model performance as compared to using oil and gas datasets that are publicly available.