Search NASA⌕ Search

SEARCH · Search NASA

Results for “Domain knowledge”

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 649 records · Page 36

Synthesis of the land carbon fluxes of the Amazon region between 2010 and 2020

The Amazon is the largest continuous tropical forest in the world and plays a key role in the global carbon cycle. Human-induced disturbances and climate change have impacted the Amazon carbon balance. Here we conduct a comprehensive synthesis of existing state-of-the-art estimates of the contemporary land carbon fluxes in the Amazon using a set of bottom-up methods (i.e., dynamic vegetation models and bookkeeping models) and a top-down inversion (atmospheric inversion model) over the Brazilian Amazon and the whole Biogeographical Amazon domain. Over the whole biogeographical Amazon region bottom-up methodologies suggest a small average carbon sink over 2010-2020, in contrast to a small carbon source simulated by top-down inversion (2010-2018). However, these estimates are not significantly different from one another when accounting for their large individual uncertainties, highlighting remaining knowledge gaps, and the urgent need to reduce such uncertainties. Nevertheless, both methodologies agreed that the Brazilian Amazon has been a net carbon source during recent climate extremes and that the south-eastern Amazon was a net land carbon source over the whole study period (2010-2020). Overall, our results point to increasing human-induced disturbances (deforestation and forest degradation by wildfires) and reduction in the old-growth forest sink during drought.

54 ENVIRONMENTAL SCIENCES↗

A Transfer Learning Strategy for Improving the Data Efficiency of Deep Reinforcement Learning Control in Smart Buildings

Reinforcement learning (RL) is a powerful tool that has shown promising results in many domains such as robotics and game-playing. Because RL algorithms learn optimal control policies by continuously interacting with their environments, these algorithms require a lot of data to learn, which limits their application to a wide range of domains. For this reason, there is an immense need for improving the training and data efficiency of RL. Towards addressing this research gap, this paper proposes a transfer learning (TL) approach to improve the efficiency of the RL algorithms by reducing data need and, thus, reducing training time. To demonstrate the proposed approach, a knowledge transfer from a set of buildings to another building was conducted. The results show that the proposed TL approach is a promising method that can efficiently harness the information from similar RL tasks and reduce the data needs of RL algorithms.

Amasyali, Kadir↗

MeDICi Software Superglue for Data Analysis Pipelines

The Middleware for Data-Intensive Computing (MeDICi) Integration Framework is an integrated middleware platform developed to solve data analysis and processing needs of scientists across many domains. MeDICi is scalable, easily modified, and robust to multiple languages, protocols, and hardware platforms, and in use today by PNNL scientists for bioinformatics, power grid failure analysis, and text analysis.

96 KNOWLEDGE MANAGEMENT AND PRESERVATION↗

Structural, functional, and genetic analyses of the actinobacterial transcription factor RbpA

Gene expression is highly regulated at the step of transcription initiation, and transcription activators play a critical role in this process. RbpA, an actinobacterial transcription activator that is essential in Mycobacterium tuberculosis (Mtb), binds selectively to group 1 and certain group 2 σ-factors. To delineate the molecular mechanism of RbpA, we show that the Mtb RbpA σ-interacting domain (SID) and basic linker are sufficient for transcription activation. We also present the crystal structure of the Mtb RbpA-SID in complex with domain 2 of the housekeeping σ-factor, σ A . The structure explains the basis of σ-selectivity by RbpA, showing that RbpA interacts with conserved regions of σ A as well as the nonconserved region (NCR), which is present only in housekeeping σ-factors. Thus, the structure is the first, to our knowledge, to show a protein interacting with the NCR of a σ-factor. We confirm the basis of selectivity and the observed interactions using mutagenesis and functional studies. In addition, the structure allows for a model of the RbpA-SID in the context of a transcription initiation complex. Unexpectedly, the structural modeling suggests that RbpA contacts the promoter DNA, and we present in vivo and in vitro studies supporting this finding. As a result, our combined data lead to a better understanding of the mechanism of RbpA function as a transcription activator.

59 BASIC BIOLOGICAL SCIENCES↗

Utilization of the Differential Die-Away Self-Interrogation Technique for Characterization and Verification of Spent Nuclear Fuel

New nondestructive assay techniques are sought to better characterize spent nuclear fuel. One of the NDA instruments selected for possible deployment is differential die-away self-interrogation (DDSI). The proposed DDSI approach for spent fuel assembly assay utilizes primarily the spontaneous fission and (α, n) neutrons in the assemblies as an internal interrogating radiation source. The neutrons released in spontaneous fission or (α,n) reactions are thermalized in the surrounding water and induce fission in fissile isotopes, thereby creating a measurable signal from isotopes of interest that would be otherwise difficult to measure. The DDSI instrument employs neutron coincidence counting with 3He tubes and list-mode-based data acquisition to allow for production of Rossi-alpha distributions (RADs) in post-processing. The list-mode approach to data collection and subsequent construction of RADs has expanded the analytical possibilities, as will be demonstrated throughout this thesis. One of the primary advantages is that the measured signal in the form of a RAD can be analyzed in its entirety including determination of die-away times in different time domains. This capability led to the development of the early die-away method, a novel leakage multiplication determination method which is tested throughout the thesis on different sources in simulation space and fresh fuel experiments. The early die-away method is a robust, accurate, improved method of determining multiplication without the need for knowledge of the (α,n) source term. The DDSI technique and instrument are presented along with the many novel capabilities enabled by and discovered through RAD analysis. Among the new capabilities presented are the early die-away method, total plutonium content determination, and highly sensitive missing pin detection. Simulation of hundreds of different spent and fresh fuel assemblies were used to develop the analysis algorithms and the techniques were tested on a variety of spontaneous fission-driven fresh fuel assemblies at Los Alamos National Laboratory and the BeRP ball at the Nevada National Security Site. The development of the new, improved analysis and characterization methods with the DDSI instrument makes it a viable technique for implementation in a facility to meet material control and safeguards needs.

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗

Effects of residual oxygen on superconducting niobium films

The integration of niobium (Nb) into emerging superconducting circuits can enhance their performance and function. However, growth of high purity Nb can be challenging due to its high reactivity with oxygen. Here, in this study, we examine the role of residual oxygen inside the growth chamber in transforming the structural, chemical, and superconducting properties of Nb films. We demonstrate that an increase in unintentional oxygen impurities lowers the superconducting critical temperature of Nb. This evolution coincides with the reduction of Nb crystal domains, which are separated by highly disordered oxygen-rich regions. Moreover, chemical analysis reveals the formation of niobium monoxide within the film during growth. These findings provide a comprehensive picture of how residual oxygen in the growth chamber can affect the properties of the Nb films. This study contributes to the materials science and engineering knowledge of superconducting Nb growth.

36 MATERIALS SCIENCE↗

Technology and Research Requirements for Combating Human Trafficking: Enhancing Communication, Analysis, Reporting, and Information Sharing

DHS’ Science & Technology Directorate directed PNNL to conduct an exploratory study on the domain of human trafficking in the Pacific Northwest in order to examine and identify technology and research requirements for enhancing communication, analysis, reporting, and information sharing – activities that directly support efforts to track, identify, deter, and prosecute human trafficking – including identification of potential national threats from smuggling and trafficking networks. This effort was conducted under the Knowledge Management Technologies Portfolio as part of the Integrated Federal, State, and Local/Regional Information Sharing (RISC) and Collaboration Program.

99 GENERAL AND MISCELLANEOUS↗

The Co-Transport of PFAS and Cr(VI) in porous media

PFAS and Cr are present at some sites as co-contaminants. Here, the objective of this research was to investigate the co-transport behavior of per- and polyfluoroalkyl substances (PFAS) and hexavalent chromium (Cr(VI)) in porous media. Miscible-displacement experiments were conducted using two soils and an aquifer sediment with different geochemical properties. Perfluorooctane sulfonic acid (PFOS) and perfluorooctanoic acid (PFOA) were employed as model PFAS. The retardation of PFOS was decreased in the presence of Cr(VI). Conversely, the transport and retardation of PFOA was not affected by the presence of Cr(VI). The reduction of PFOS retardation caused by Cr(VI) is likely due to sorption competition for both organic-carbon and inorganic (metal-oxides and clay minerals) domains. The relative contributions of the three soil constituents to PFOS sorption and the potential for competition between PFOS and Cr(VI) is a function of the geochemical composition of the porous media (i.e., organic carbon, metal-oxides and clay minerals). The PFAS had minimal impact on the retention and transport of Cr(VI). To our knowledge, the results presented herein represent the first reported data for PFOS and Cr(VI) co-transport in porous media. The results of this study indicate that the presence of Cr(VI) has the potential to increase the migration potential of PFOS in soil and groundwater, which should be considered when characterizing electroplating facilities, leather tanning facilities, and other co-contaminated sites.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Quantifying Mixed Uncertainties in Cyber Attacker Payoffs

Representation and propagation of uncertainty in cyber attacker payoffs is a key aspect of security games. Past research has primarily focused on representing the defender’s beliefs about attacker payoffs as point utility estimates. More recently, within the physical security domain, attacker payoff uncertainties have been represented as Uniform and Gaussian probability distributions, and intervals. Within cyber-settings, continuous probability distributions may still be appropriate for addressing statistical (aleatory) uncertainties where the defender may assume that the attacker’s payoffs differ over time. However, systematic (epistemic) uncertainties may exist, where the defender may not have sufficient knowledge or there is insufficient information about the attacker’s payoff generation mechanism. Such epistemic uncertainties are more suitably represented as probability boxes with intervals. In this study, we explore the mathematical treatment of such mixed payoff uncertainties.

Chatterjee, Samrat↗

Dynamic interactions of proteins in complex networks

Recent advances in techniques such as NMR and EPR spectroscopy have enabled the elucidation of how proteins undergo structural changes to act in concert in complex networks. The three minireviews in this series highlight current findings and the capabilities of new methodologies for unraveling the dynamic changes controlling diverse cellular functions. They represent a sampling of the cutting-edge research presented at the 17th Meeting of Methods in Protein Structure Analysis, MPSA2008, in Sapporo, Japan, 26-29 August, 2008 (http://www.iapsap.bnl.gov). The first minireview, by Christensen and Klevit, reports on a structure-based yeast two-hybrid method for identifying E2 ubiquitin-conjugating enzymes that interact with the E3 BRCA1/BARD1 heterodimer ligase to generate either mono- or polyubiquitinated products. This method demonstrated for the first time that the BRCA1/BARD1 E3 can interact with 10 different E2 enzymes. Interestingly, the interaction with multiple E2 enzymes displayed unique ubiquitin-transfer properties, a feature expected to be common among other RING and U-box E3s. Further characterization of new E3 ligases and the E2 enzymes that interact with them will greatly enhance our understanding of ubiquitin transfer and facilitate studies of roles of ubiquitin and ubiquitin-like proteins in protein processing and trafficking. Stein et al., in the second minireview, describe recent progress in defining the binding specificity of different peptide-binding domains. The authors clearly point out that transient peptide interactions mediated by both post-translational modifications and disordered regions ensure a high level of specificity. They postulate that a regulatory code may dictate the number of combinations of domains and post-translational modifications needed to achieve the required level of interaction specificity. Moreover, recognition alone is not enough to obtain a stable complex, especially in a complex cellular environment. Increasing evidence indicates that disordered domains can acquire structural features that modulate the binding and strength of the signaling cascade. Whereas the first two minireviews describe ways in which protein interactions are modulated, the third, by Tompa, focuses on the importance of protein disorder in a subset of amyloid proteins. It is apparent that within this group, part of the polypeptide chain remains disordered during amyloid formation. Moreover, the disordered segments have different amino acid composition and physicochemical characteristics, which suggests that they may play a role in amyloid stability. The disordered region may serve as a linker to connect the ordered core and a globular domain, maintaining the stability and structure of the globular domain and minimizing protein refolding upon amyloid formation. As techniques in protein chemistry advance, we are learning more and more about the mechanisms that regulate and are regulated by protein interactions. The three minireviews in this series offer a glimpse of the complex dynamics fundamental to protein-protein interactions. In the future, we expect that the knowledge gained will help to augment our ability to control complex pathologies and treat diverse diseases states.

59 BASIC BIOLOGICAL SCIENCES↗

SU-E-T-51: Bayesian Network Models for Radiotherapy Error Detection

Purpose: To develop a probabilistic model of radiotherapy plans using Bayesian networks that will detect potential errors in radiation delivery. Methods: Semi-structured interviews with medical physicists and other domain experts were employed to generate a set of layered nodes and arcs forming a Bayesian Network (BN) which encapsulates relevant radiotherapy concepts and their associated interdependencies. Concepts in the final network were limited to those whose parameters are represented in the institutional database at a level significant enough to develop mathematical distributions. The concept-relation knowledge base was constructed using the Web Ontology Language (OWL) and translated into Hugin Expert Bayes Network files via the the RHugin package in the R statistical programming language. A subset of de-identified data derived from a Mosaiq relational database representing 1937 unique prescription cases was processed and pre-screened for errors and then used by the Hugin implementation of the Estimation-Maximization (EM) algorithm for machine learning all parameter distributions. Individual networks were generated for each of several commonly treated anatomic regions identified by ICD-9 neoplasm categories including lung, brain, lymphoma, and female breast. Results: The resulting Bayesian networks represent a large part of the probabilistic knowledge inherent in treatment planning. By populating the networks entirely with data captured from a clinical oncology information management system over the course of several years of normal practice, we were able to create accurate probability tables with no additional time spent by experts or clinicians. These probabilistic descriptions of the treatment planning allow one to check if a treatment plan is within the normal scope of practice, given some initial set of clinical evidence and thereby detect for potential outliers to be flagged for further investigation. Conclusion: The networks developed here support the use of probabilistic models into clinical chart checking for improved detection of potential errors in RT plans.

60 APPLIED LIFE SCIENCES↗

Final Report - Phylogenomic tools and web resources for the Systems Biology Knowledgebase

The major advance during this last reporting period (8/15/12 to present) is our release of data on the PhyloFacts website: phylogenetic trees, multiple sequence alignments and other data for protein families are now available for download from http://phylogenomics.berkeley.edu/data/. This project as a whole aimed to develop high-throughput functional annotation systems that exploit information from protein 3D structure and evolution to provide highly precise inferences of various aspects of gene function, including molecular function, biological process, pathway association, Pfam domains, cellular localization and so on. We accomplished these aims by developing and testing different systems on a database of protein family trees: the PhyloFacts Phylogenomic Encyclopedia (at http://phylogenomics.berkeley.edu/phylofacts/ ).

60 APPLIED LIFE SCIENCES↗

Coordinated Unmanned Aircraft System (UAS) and Ground-Based Weather Measurements to Predict Lagrangian Coherent Structures (LCSs)

Concentrations of airborne chemical and biological agents from a hazardous release are not spread uniformly. Instead, there are regions of higher concentration, in part due to local atmospheric flow conditions which can attract agents. We equipped a ground station and two rotary-wing unmanned aircraft systems (UASs) with ultrasonic anemometers. Flights reported here were conducted 10 to 15 m above ground level (AGL) at the Leach Airfield in the San Luis Valley, Colorado as part of the Lower Atmospheric Process Studies at Elevation—a Remotely-Piloted Aircraft Team Experiment (LAPSE-RATE) campaign in 2018. The ultrasonic anemometers were used to collect simultaneous measurements of wind speed, wind direction, and temperature in a fixed triangle pattern; each sensor was located at one apex of a triangle with ∼100 to 200 m on each side, depending on the experiment. A WRF-LES model was used to determine the wind field across the sampling domain. Data from the ground-based sensors and the two UASs were used to detect attracting regions (also known as Lagrangian Coherent Structures, or LCSs), which have the potential to transport high concentrations of agents. This unique framework for detection of high concentration regions is based on estimates of the horizontal wind gradient tensor. To our knowledge, our work represents the first direct measurement of an LCS indicator in the atmosphere using a team of sensors. Our ultimate goal is to use environmental data from swarms of sensors to drive transport models of hazardous agents that can lead to real-time proper decisions regarding rapid emergency responses. The integration of real-time data from unmanned assets, advanced mathematical techniques for transport analysis, and predictive models can help assist in emergency response decisions in the future.

54 ENVIRONMENTAL SCIENCES↗

A Probabilistic Framework for Quantifying Mixed Uncertainties in Cyber Attacker Payoffs

Quantification and propagation of uncertainties in cyber attacker payoffs is a key aspect within multiplayer, stochastic security games. These payoffs may represent penalties or rewards associated with player actions and are subject to various sources of uncertainty, including: (1) cyber-system state, (2) attacker type, (3) choice of player actions, and (4) cyber-system state transitions over time. Past research has primarily focused on representing defender beliefs about attacker payoffs as point utility estimates. More recently, within the physical security domain, attacker payoff uncertainties have been represented as Uniform and Gaussian probability distributions, and mathematical intervals. For cyber-systems, probability distributions may help address statistical (aleatory) uncertainties where the defender may assume inherent variability or randomness in the factors contributing to the attacker payoffs. However, systematic (epistemic) uncertainties may exist, where the defender may not have sufficient knowledge or there is insufficient information about the attacker’s payoff generation mechanism. Such epistemic uncertainties are more suitably represented as generalizations of probability boxes. This paper explores the mathematical treatment of such mixed payoff uncertainties. A conditional probabilistic reasoning approach is adopted to organize the dependencies between a cyber-system’s state, attacker type, player actions, and state transitions. This also enables the application of probabilistic theories to propagate various uncertainties in the attacker payoffs. An example implementation of this probabilistic framework and resulting attacker payoff distributions are discussed. A goal of this paper is also to highlight this uncertainty quantification problem space to the cyber security research community and encourage further advancements in this area.

cyber attacker payoffs↗

On the universal structure of human lexical semantics

How universal is human conceptual structure? The way concepts are organized in the human brain may reflect distinct features of cultural, historical, and environmental background in addition to properties universal to human cognition. Semantics, or meaning expressed through language, provides indirect access to the underlying conceptual structure, but meaning is notoriously difficult to measure, let alone parameterize. Here, we provide an empirical measure of semantic proximity between concepts using cross-linguistic dictionaries to translate words to and from languages carefully selected to be representative of worldwide diversity. These translations reveal cases where a particular language uses a single “polysemous” word to express multiple concepts that another language represents using distinct words. We use the frequency of such polysemies linking two concepts as a measure of their semantic proximity and represent the pattern of these linkages by a weighted network. This network is highly structured: Certain concepts are far more prone to polysemy than others, and naturally interpretable clusters of closely related concepts emerge. Statistical analysis of the polysemies observed in a subset of the basic vocabulary shows that these structural properties are consistent across different language groups, and largely independent of geography, environment, and the presence or absence of a literary tradition. As a result, the methods developed here can be applied to any semantic domain to reveal the extent to which its conceptual structure is, similarly, a universal attribute of human cognition and language use.

96 KNOWLEDGE MANAGEMENT AND PRESERVATION↗

Advancing equitable health and well-being across urban–rural sustainable infrastructure systems

Infrastructure systems have direct implications for how health and well-being evolve across urban–rural systems. Scientists, practitioners, and policy-makers use domain-specific methods and tools to characterize sectors of infrastructure, but these approaches do not capture the cascading effects across interrelated infrastructure and governance domains. We argue that the development and management of sustainable urban infrastructure must focus on interactions across urban and rural places to advance equitable health and well-being. We call for a research agenda that focuses on urban–rural infrastructure systems, addressing trade-offs and synergies, decision-making, institutional arrangements, and effective co-production of knowledge across the diverse places connected by infrastructure.

60 APPLIED LIFE SCIENCES↗

Application of PCM Validation to BFBT Experiments

In support of model validation, this summary applies a recently proposed methodology, denoted hereinafter by physics-guided coverage mapping (PCM), to the estimation of void fraction in heated channels using the OECD/NEA full size fine mesh bundle test (BFBT) data. PCM here is developed as a quantitative approach that allows for a rigorous scaling of experimental biases to the application domain, i.e., the range of the conditions intended for model application. PCM aims to address some of the conceptual and computational difficulties associated with model validation exercises where it is essential to establish with a certain confidence that the available experiments provide enough evidence to support the use of the model for its intended range of application. To our knowledge, this methodology represents a first-of-a-kind application of physics models to support validation, in a manner that precludes parameters calibration, and avoids the introduction of expert-based scale-distortion uncertainties. (authors)

42 ENGINEERING↗

Reinforcement Learning for feedback-enabled cyber resilience

The rapid growth in the number of devices and their connectivity has enlarged the attack surface and made cyber systems more vulnerable. As attackers become increasingly sophisticated and resourceful, mere reliance on traditional cyber protection, such as intrusion detection, firewalls, and encryption, is insufficient to secure the cyber systems. Cyber resilience provides a new security paradigm that complements inadequate protection with resilience mechanisms. A Cyber-Resilient Mechanism (CRM) adapts to the known or zero-day threats and uncertainties in real-time and strategically responds to them to maintain the critical functions of the cyber systems in the event of successful attacks. Feedback architectures play a pivotal role in enabling the online sensing, reasoning, and actuation process of the CRM. Reinforcement Learning (RL) is an important gathering of algorithms that epitomize the feedback architectures for cyber resilience. It allows the CRM to provide dynamic and sequential responses to attacks with limited or without prior knowledge of the environment and the attacker. In this work, we review the literature on RL for cyber resilience and discuss the cyber-resilient defenses against three major types of vulnerabilities, i.e., posture-related, information-related, and human-related vulnerabilities. Here we introduce moving target defense, defensive cyber deception, and assistive human security technologies as three application domains of CRMs to elaborate on their designs. The RL algorithms also have vulnerabilities themselves. We explain the major vulnerabilities of RL and present develop several attack models where the attacker target the information exchanged between the environment and the agent: the rewards, the state observations, and the action commands. We show that the attacker can trick the RL agent into learning a nefarious policy with minimum attacking effort. The paper introduces several defense methods to secure the RL-enabled systems from these attacks. However, there is still a lack of works that focuses on the defensive mechanisms for RL-enabled systems. Last but not least, we discuss the future challenges of RL for cyber security and resilience and emerging applications of RL-based CRMs.

97 MATHEMATICS AND COMPUTING↗