Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “training data quality”

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 109 records · Page 6
All sourcesNASA NTRSDOE OSTI
DOE OSTI2024

Harnessing Machine Learning and Data Fusion for Accurate Undocumented Well Identification in Satellite Images

This study utilizes satellite data to detect undocumented oil and gas wells, which pose significant environmental concerns, including greenhouse gas emissions. Three key findings emerge from the study. Firstly, the problem of imbalanced data is addressed by recommending oversampling techniques like Rotation–GaussianBlur–Solarization data augmentation (RGS), the Synthetic Minority Over-Sampling Technique (SMOTE), or ADASYN (an extension of SMOTE) over undersampling techniques. The performance of borderline SMOTE is less effective than that of the rest of the oversampling techniques, as its performance relies heavily on the quality and distribution of data near the decision boundary. Secondly, incorporating pre-trained models trained on large-scale datasets enhances the models’ generalization ability, with models trained on one county’s dataset demonstrating high overall accuracy, recall, and F1 scores that can be extended to other areas. This transferability of models allows for wider application. Lastly, including persistent homology (PH) as an additional input improves performance for in-distribution testing but may affect the model’s generalization for out-of-distribution testing. A careful consideration of PH’s impact on overall performance and generalizability is recommended. Overall, this study provides a robust approach to identifying undocumented oil and gas wells, contributing to the acceleration of a net-zero economy and supporting environmental sustainability efforts.

SMOTE↗
DOE OSTI2023

Physics-Driven Convolutional Autoencoder Approach for CFD Data Compressions: Preprint

With the growing size and complexity of turbulent flow models, data compression approaches are of the utmost importance to analyze, visualize, or restart the simulations. Recently, in-situ autoencoder-based compression approaches have been proposed and shown to be effective at producing reduced representations of turbulent flow data. However, these approaches focus solely on training the model using point-wise sample reconstruction losses that do not take advantage of the physical properties of turbulent flows. In this paper, we show that training autoencoders with additional physics-informed regularizations, e.g., enforcing incompressibility and preserving enstrophy, improves the compression model in three ways: (i) the compressed data better conform to known physics for homogeneous isotropic turbulence without negatively impacting point-wise reconstruction quality, (ii) inspection of the gradients of the trained model uncovers changes to the learned compression mapping that can facilitate the use of explainability techniques, and (iii) as a performance byproduct, training losses are shown to converge up to 12x faster than the baseline model.

auto-encoders↗
DOE OSTI2025

Automated segmentation and analysis of point clouds of pier foundations using Pier Inspection and Evaluation Report (PIER)

Pier foundations are commonly used in locations with unstable soil or where other types of foundations are unsuitable or cost prohibitive. A pier foundation consists of vertical columns to support the structure and elevate it above the ground. Common materials for pier foundations include masonry, concrete, timber, and steel. The methods for accurate placement of pier foundations have remained relatively unchanged for decades. For simple installations, construction chalk lines are used to layout the locations of piers to ensure accurate placement and elevation. For more complex installations, surveying instruments operated by trained professionals are employed to accurately locate piers and assess correct elevation before construction. After installation, another survey may need to be performed to assess the quality of the as-built foundation. However, with the advent of terrestrial laser scanners (TLS), the means now exist for contractors to conduct their own assessments of as-built foundations. The major barrier preventing contractors from performing their own assessments of as-built foundation quality is the segmentation and analysis of point cloud data, a skill that often requires a trained user. The objective of this research is to develop a software tool (PIER: Pier Inspection and Evaluation Report) to enable automated segmentation and analysis of point clouds of pier foundations. In this paper, the automated segmentation and analysis algorithms are detailed. A mockup lay out of pier foundations was built using concrete masonry units, and the algorithms were tested to evaluate performance. Limitations of the current algorithms and future research direction are discussed.

Turki, Amine [ORNL]↗
DOE OSTI2021

Characterizing manufacturing wastewater in the United States for the purpose of analyzing energy requirements for reuse

This paper seeks to inform an improved understanding of the energy tradeoff associated with on-site manufacturing water reuse in the United States from a lifecycle perspective, in part by developing an analytical framework for understanding when this tradeoff for reuse is beneficial. We survey the literature to assess the current state of reuse and its motives and barriers in the United States, before synthesizing information from publicly available EPA data on contaminants in US manufacturing wastewaters and technologies for treating them. Using the available data, we derive a set of “ubiquitous contaminants” among the top ten in terms of mass discharged in more than half of US manufacturing subsectors (NAICS 31–33) according to EPA permit data. We also present information on proven treatment trains and their energy requirements. We then compare water quality requirements for specific contaminants in reclaimed water to those characteristic of wastewater streams currently being discharged from manufacturing plants into surface waters to highlight sectors with reuse opportunities that could require little cost to realize, such as primary metals and, to a lesser extent, petroleum and coal products. We conclude by highlighting data limitations that need to be rectified before applying the framework more broadly and discussing how these data gaps could be filled. Better understanding the relationship between energy and water in the context of on-site manufacturing water reuse would allow manufacturers to improve resiliency by reducing regulatory, physical, and reputational risks while lessening their footprint on local watersheds.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗
DOE OSTI2022

Preliminary Results for Using Uncertainty and Out-of-distribution Detection to Identify Unreliable Predictions.

As machine learning (ML) models are deployed into an ever-diversifying set of application spaces, ranging from self-driving cars to cybersecurity to climate modeling, the need to carefully evaluate model credibility becomes increasingly important. Uncertainty quantification (UQ) provides important information about the ability of a learned model to make sound predictions, often with respect to individual test cases. However, most UQ methods for ML are themselves data-driven and therefore susceptible to the same knowledge gaps as the models themselves. Specifically, UQ helps to identify points near decision boundaries where the models fit the data poorly, yet predictions can score as certain for points that are under-represented by the training data and thus out-of-distribution (OOD). One method for evaluating the quality of both ML models and their associated uncertainty estimates is out-of-distribution detection (OODD). We combine OODD with UQ to provide insights into the reliability of the individual predictions made by an ML model.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2020

Real-time monitoring and prediction of water quality parameters and algae concentrations using microbial potentiometric sensor signals and machine learning tools

We report the overarching hypothesis of this study was that temporal microbial potentiometric sensor (MPS) signal patterns could be used to predict changes in commonly monitored water quality parameters by using artificial intelligence/machine learning tools. To test this hypothesis, the study first examines a proof of concept by correlating between MPS's signals and high algae concentrations in an algal cultivation pond. Then, the study expanded upon these findings and examined if multiple water quality parameters could be predicted in real surface waters, like irrigation canals. Signals generated between the MPS sensors and other water quality sensors maintained by an Arizona utility company, including algae and chlorophyll, were collected in real time at time intervals of 30 min over a period of 9 months. Data from the MPS system and data collected by the utility company were used to train the ML/AI algorithms and compare the predicted with actual water quality parameters and algae concentrations. Based on the composite signal obtained from the MPS, the ML/AI was used to predict the canal surface water's turbidity, conductivity, chlorophyll, and blue-green algae (BGA), dissolved oxygen (DO), and pH, and predicted values were compared to the measured values. Initial testing in the algal cultivation pond revealed a strong linear correlation (R 2 = 0.87) between mixed liquor suspended solids (MLSS) and the MPSs' composite signals. The Normalized Root Mean Square Error (NRMSE) between the predicted values and measured values were <6.5%, except for the DO, which was 10.45%. The results demonstrate the usefulness of MPSs to predict key surface water quality parameters through a single composite signal, when the ML/AI tools are used conjunctively to disaggregate these signal components. The maintenance-free MPS offers a novel and cost-effective approach to monitor numerous water quality parameters at once with relatively high accuracy.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2023

DOLCE: A Model-Based Probabilistic Diffusion Framework for Limited-Angle CT Reconstruction

Limited-Angle Computed Tomography (LACT) is a non-destructive evaluation technique used in a variety of applications ranging from security to medicine. The limited angle coverage in LACT is often a dominant source of severe artifacts in the reconstructed images, making it a challenging inverse problem. We present DOLCE, a new deep model-based framework for LACT that uses a conditional diffusion model as an image prior. Diffusion models are a recent class of deep generative models that are relatively easy to train due to their implementation as image denoisers. DOLCE can form high-quality images from severely under-sampled data by integrating data-consistency updates with the sampling updates of a diffusion model, which is conditioned on the transformed limited-angle data. We show through extensive experimentation on several challenging real LACT datasets that, the same pre-trained DOLCE model achieves the SOTA performance on drastically different types of images. Additionally, we show that, unlike standard LACT reconstruction methods, DOLCE naturally enables the quantification of the reconstruction uncertainty by generating multiple samples consistent with the measured data.

Kim, Hyojin↗
DOE OSTI2021

A Comparison of Machine Learning Methods to Forecast Tropospheric Ozone Levels in Delhi

Ground-level ozone is a pollutant that is harmful to urban populations, particularly in developing countries where it is present in significant quantities. It greatly increases the risk of heart and lung diseases and harms agricultural crops. This study hypothesized that, as a secondary pollutant, ground-level ozone is amenable to 24 h forecasting based on measurements of weather conditions and primary pollutants such as nitrogen oxides and volatile organic compounds. We developed software to analyze hourly records of 12 air pollutants and 5 weather variables over the course of one year in Delhi, India. To determine the best predictive model, eight machine learning algorithms were tuned, trained, tested, and compared using cross-validation with hourly data for a full year. The algorithms, ranked by R2 values, were XGBoost (0.61), Random Forest (0.61), K-Nearest Neighbor Regression (0.55), Support Vector Regression (0.48), Decision Trees (0.43), AdaBoost (0.39), and linear regression (0.39). When trained by separate seasons across five years, the predictive capabilities of all models increased, with a maximum R 2 of 0.75 during winter. Bidirectional Long Short-Term Memory was the least accurate model for annual training, but had some of the best predictions for seasonal training. Out of five air quality index categories, the XGBoost model was able to predict the correct category 24 h in advance 90% of the time when trained with full-year data. Separated by season, winter is considerably more predictable (97.3%), followed by post-monsoon (92.8%), monsoon (90.3%), and summer (88.9%). These results show the importance of training machine learning methods with season-specific data sets and comparing a large number of methods for specific applications.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2020

Multimodal Data Fusion via Entropy Minimization

The use of gradient-based data-driven models to solve a range of real-world remote sensing problems can in practice be limited by the uniformity of available data. Use of data from disparate sensor types, resolutions, and qualities typically requires compromises based on assumptions that are made prior to model training and may not necessarily be optimal given over-arching objectives. For example, while deep neural networks (NNs) are state-of-the-art in a variety of target detection problems, training them typically requires either limiting the training data to a subset over which uniformity can be enforced or training independent models which subsequently require additional score fusion. The method we introduce here seeks to leverage the benefits of both approaches by allowing correlated inputs from different data sources to co-influence preferred model solutions, while maintaining flexibility over missing and mismatching data. In this work we propose a new data fusion technique for gradient updated models based on entropy minimization and experimentally validate it on a hyperspectral target detection dataset. We demonstrate superior performance compared to currently available techniques using a range of realistic data scenarios, where available data has limited spacial overlap and resolution.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2023

Temperature outweighs light and flow as the predominant driver of dissolved oxygen in US rivers

The concentration of dissolved oxygen (DO), an important measure of water quality and river metabolism, varies tremendously in time and space. Riverine DO is commonly perceived as regulated by interacting and competing drivers (light, temperature and flow) that define rivers’ climate. Its continental-scale drivers, however, have remained elusive, partly due to the scarcity and spatio-temporal inconsistency of water quality data. Here we show, via a deep learning model (long short-term memory) trained using data from 580 rivers, that temperature predominantly drives daily DO dynamics in the contiguous United States. Light comes a close second, whereas flow imparts minimal influence. This work showcases the promise of using deep learning models for data filling that enables large-scale systematic analysis of patterns and drivers. Results show fairly accurate prediction of DO by temperature alone, and declining DO in warming rivers, which has important implications for water security and ecosystem health in the future climate.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2021

Scalable balanced training of conditional generative adversarial neural networks on image data

Here, we propose a distributed approach to train deep convolutional generative adversarial neural network (DC-CGANs) models. Our method reduces the imbalance between generator and discriminator by partitioning the training data according to data labels, and enhances scalability by performing a parallel training where multiple generators are concurrently trained, each one of them focusing on a single data label. Performance is assessed in terms of inception score, Fréchet inception distance, and image quality on MNIST, CIFAR10, CIFAR100, and ImageNet1k datasets, showing a significant improvement in comparison to state-of-the-art techniques to training DC-CGANs. Weak scaling is attained on all the four datasets using up to 1000 processes and 2000 NVIDIA V100 GPUs on the OLCF supercomputer Summit.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2021

Direct Prediction of Phonon Density of States With Euclidean Neural Networks

Abstract Machine learning has demonstrated great power in materials design, discovery, and property prediction. However, despite the success of machine learning in predicting discrete properties, challenges remain for continuous property prediction. The challenge is aggravated in crystalline solids due to crystallographic symmetry considerations and data scarcity. Here, the direct prediction of phonon density‐of‐states (DOS) is demonstrated using only atomic species and positions as input. Euclidean neural networks are applied, which by construction are equivariant to 3D rotations, translations, and inversion and thereby capture full crystal symmetry, and achieve high‐quality prediction using a small training set of examples with over 64 atom types. The predictive model reproduces key features of experimental data and even generalizes to materials with unseen elements, and is naturally suited to efficiently predict alloy systems without additional computational cost. The potential of the network is demonstrated by predicting a broad number of high phononic specific heat capacity materials. The work indicates an efficient approach to explore materials' phonon structure, and can further enable rapid screening for high‐performance thermal storage materials and phonon‐mediated superconductors.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2020

Autonomous Tuning and Charge-State Detection of Gate-Defined Quantum Dots

Defining quantum dots in semiconductor-based heterostructures is an essential step in initializing solid-state qubits. With growing device complexity and increasing number of functional devices required for measurements, a manual approach to finding suitable gate voltages to confine electrons electrostatically is impractical. Here, we implement a two-stage device characterization and dot-tuning process, which first determines whether devices are functional and then attempts to tune the functional devices to the single or double quantum-dot regime. We show that automating well-established manual-tuning procedures and replacing the experimenter’s decisions by supervised machine learning is sufficient to tune double quantum dots in multiple devices without premeasured input or manual intervention. The quality of measurement results and charge states are assessed by four binary classifiers trained with experimental data, reflecting real device behavior. We compare and optimize eight models and different data preprocessing techniques for each of the classifiers to achieve reliable autonomous tuning, an essential step towards scalable quantum systems in quantum-dot-based qubit architectures.

Darulova, J↗
DOE OSTI2022

Simurgh: A Framework for Cad-Driven Deep Learning Based X-Ray CT Reconstruction

High-resolution X-ray computed tomography (XCT) is an important technique for the inspection of additively manufactured (AM) parts. While XCT is typically used off-line to inspect a subset of manufactured parts, significantly accelerating measurement speed while retaining accuracy would enable use of XCT for in-line inspection to rapidly identify defects in each part as it is manufactured. Here, we propose a deep learning (DL) based approach that uses computer aided design (CAD) models of the AM parts and physics-based information to rapidly produce high-quality reconstructions from sparse XCT measurements without high quality ground truth data. Our approach uses a generative adversarial neural network (GAN) to produced realistic training data from the CAD-based simulations and a deep neural network that is trained using data from the first stage to produce accurate 3D reconstructions. Using experimental XCT data of metal parts, we demonstrate enhanced defect detection capabilities while dramatically reducing the scan time.

Ziabari, Amir↗
DOE OSTI2024

RAP: Resource-aware Automated GPU Sharing for Multi-GPU Recommendation Model Training and Input Preprocessing

Ensuring high-quality recommendations for newly onboarded users requires the continuous retraining of Deep Learning Recommendation Models (DLRMs) with freshly generated data. To serve the online DLRM retraining, existing solutions use hundreds of CPU computing nodes designated for input preprocessing, causing significant power consumption that surpasses even the power usage of GPU trainers. To this end, we propose RAP, an end-to-end DLRM training framework that supports Resource-aware Automated GPU sharing for DLRM input Preprocessing and Training. The core idea of RAP is to accurately capture the remaining GPU computing resources during DLRM training for input preprocessing, achieving superior training efficiency without requiring additional resources. Specifically, RAP utilizes a co-running cost model to efficiently assess the costs of various input preprocessing operations, and it implements a resource-aware horizontal fusion technique that adaptively merges smaller kernels according to GPU availability, circumventing any interference with DLRM training. In addition, RAP leverages a heuristic searching algorithm that jointly optimizes both the input preprocessing graph mapping and the co-running schedule to maximize the end-to-end DLRM training throughput. The comprehensive evaluation shows that RAP achieves 78.3× speedup on average over CPU-based DLRM input preprocessing frameworks. In addition, the end-to-end training throughput of RAP is only 2.04% lower than the ideal case, which has no input preprocessing overhead.

Wang, Zheng↗
DOE OSTI2023

classLog: Logistic regression for the classification of genetic sequences

Introduction Sequencing and phylogenetic classification have become a common task in human and animal diagnostic laboratories. It is routine to sequence pathogens to identify genetic variations of diagnostic significance and to use these data in realtime genomic contact tracing and surveillance. Under this paradigm, unprecedented volumes of data are generated that require rapid analysis to provide meaningful inference. Methods We present a machine learning logistic regression pipeline that can assign classifications to genetic sequence data. The pipeline implements an intuitive and customizable approach to developing a trained prediction model that runs in linear time complexity, generating accurate output rapidly, even with incomplete data. Our approach was benchmarked against porcine respiratory and reproductive syndrome virus (PRRSv) and swine H1 influenza A virus (IAV) datasets. Trained classifiers were tested against sequences and simulated datasets that artificially degraded sequence quality at 0, 10, 20, 30, and 40%. Results When applied to a poor-quality sequence data, the classifier achieved between >85% to 95% accuracy for the PRRSv and the swine H1 IAV HA dataset and this increased to near perfect accuracy when using the full dataset. The model also identifies amino acid positions used to determine genetic clade identity through a feature selection ranking within the model. These positions can be mapped onto a maximum-likelihood phylogenetic tree, allowing for the inference of clade defining mutations. Discussion Our approach is implemented as a python package with code available at https://github.com/flu-crew/classLog .

Zeller, Michael A.↗
DOE OSTI2020

Real-time sparse-sampled Ptychographic imaging through deep neural networks

Ptychography has rapidly grown in the fields of X-ray and electron imaging for its unprecedented ability to achieve nano or atomic scale resolution while simultaneously retrieving chemical or magnetic information from a sample. A ptychographic reconstruction is achieved by means of solving a complex inverse problem that imposes constraints both on the acquisition and on the analysis of the data, which typically precludes real-time imaging due to computational cost involved in solving this inverse problem. In this work we propose PtychoNN, a novel approach to solve the ptychography reconstruction problem based on deep convolutional neural networks. We demonstrate how the proposed method can be used to predict real-space structure and phase at each scan point solely from the corresponding far-field diffraction data. The presented results demonstrate how PtychoNN can effectively be used on experimental data, being able to generate high quality reconstructions of a sample up to hundreds of times faster than state-of-the-art ptychography reconstruction solutions once trained. By surpassing the typical constraints of iterative model-based methods, we can significantly relax the data acquisition sampling conditions and produce equally satisfactory reconstructions. Besides drastically accelerating acquisition and analysis, this capability can enable new imaging scenarios that were not possible before, in cases of dose sensitive, dynamic and extremely voluminous samples.

Cherukara, Mathew J↗
DOE OSTI2022

Foundation Models of Scientific Knowledge for Chemistry: Opportunities, Challenges and Lessons Learned

Foundation models pre-trained on large corpora demonstrate significant gains across many natural language processing tasks and domains e.g., law, healthcare, education, etc. However, only limited efforts have investigated the opportunities and limitations of applying these powerful models to science and security applications. In this work we develop foundation models of scientific knowledge for chemistry to augment scientists with the advanced ability to perceive and reason at scale previously unimagined. Specifically, we build large-scale (1.47B parameter) general-purpose models for chemistry that can be effectively used to perform a wide range of in-domain and out-of-domain tasks. Evaluating these models in a zero-shot setting, we analyze the effect of model and data scaling, knowledge depth, and temporality on model performance in context of model training efficiency. Our novel findings demonstrate that (1) model size significantly contributes to the task performance when evaluated in a zero-shot setting; (2) data quality (aka diversity) affects model performance more than data quantity; (3) similarly, unlike previous work (Luu et al., 2021) temporal order of the documents in the corpus boosts model performance only for specific tasks, e.g., SciQ; and (4) models pre-trained from scratch perform better on in-domain tasks than those tuned from general-purpose models like Open AI’s GPT-2.

Foundation Models, Chemistry↗