SEARCH · Engineering Papers

Results for “latent space”

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.

SourceResearch connectionFrom yearThrough year

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 19 records

All sources NASA NTRS DOE OSTI

NASA NTRSDate not supplied

Use of Design of Experiments and Rule-Based Inference in Determining Neural Network Architectures for Loss of Control Detection

In this work, we describe methods for selecting the neural network architectures and input spaces to implement belief state inference on generic commercial transport aircraft. First, we highlight a case study on the planning, execution, and analysis of a set of experiments to determine the configurations of a conditional variational autoencoder (CVAE). We present a structured method that can be used in a number of aerospace applications, to optimize the structure and training parameters of the CVAE for belief state inference, using Design of Experiments (DOE) statistical methodologies. The motivation for this specific DOE was to identify the appropriate hyperparameters for measuring the CVAE reconstruction probability and latent space, such that the measurements can be used to infer qualitative state changes for the aircraft. We demonstrate that this process yields information about a trained neural network’s utility for this specific application, along with a quantifiable range of certainty. We execute 84 experiments using loss-of-control flight maneuver data from a NASA T-2 aircraft, demonstrating that this empirical process allows us to construct cheap and simple models with specific attributes amenable to belief state inference in aerospace applications. While theoretically, we could create a single CVAE with an input space the size of all measurable flight variables and environmental dynamics, it becomes intractable to use such a neural network in an in-situ intelligent multi-agent system. Using the recommendations from our case study, we introduce a technical approach for feasibly describing the belief space by (1) identifying significant statistical relationships among flight variables using rule induction, (2) using a set of rules that cover all features to define the input space of multiple CVAEs, and (3) forming a belief space based on the joint probability density of their collective latent spaces. This results in a series of relatively small matrix multiplications that can be performed in real time, as opposed to large matrix computations in a single CVAE. We demonstrate the application of this approach on the T-2 flight loss-of control experiments, using the architecture and hyperparameter recommendations from the case study. We compare the utilities of an individual CVAE trained on all flight variables and multiple CVAEs defined on subsets of flight variables for detecting qualitative changes in flight. We demonstrate that the use of multiple CVAEs with smaller input spaces permits the CVAE to capture more granular relationships in the latent space, permitting better state space characterization and loss-of-control detection.

Design of experiments↗

NASA NTRS1999

Reactivation of Latent Viruses in Space

Reactivation of latent viruses is an important health risk for people working and living in physically isolated extreme environments such as Antarctica and space. Preflight quarantine does not significantly reduce the risk associated with latent viruses, however, pharmaceutical countermeasures are available for some viruses. The molecular basis of latency is not fully understood, but physical and psychosocial stresses are known to initiate the reactivation of latent viruses. Presumably, stress induced changes in selected hormones lead to alterations in the cell- mediated immune (CMI) response resulting in increased shedding of latent viruses. Limited access to space makes the use of ground-based analogs essential. The Australian Antarctic stations serve as a good stress model and simulate many aspects of space flight. Closed environmental chambers have been used to simulate space flight since the Skylab missions and have also proven to be a valuable analog of selected aspects of space flight.

Pierson, D. L.↗

NASA NTRSDate not supplied

Loss of Control Detection for Commercial Transport Aircraft Using Conditional Variational Autoencoders

This work describes a detector for the loss of control condition of a commercial transport in flight. The detector has a belief state defined by the latent variable stochastic modeling of a conditional variational autoencoder (CVAE) constructed with bidirectional recurrent layers. In 2000, the Boeing Company and the NASA Langley Research Center jointly developed a quantitative set of metrics for defining loss-of-control (LOC) for a commercial transport. We use the thresholds for these quantitative metrics to define a condition vector for training the CVAE. First, we demonstrate through experimentation that reconstruction probability is an accurate indicator that the vehicle has shifted to an LOC state. Second, we introduce a technique for inferring that the vehicle is approaching a flight state change by measuring a shift in the sampling distributions of the CVAE latent space. The sampling distributions for flight observations that are approaching envelope limits are localized to external areas of the latent space. We provide an analysis of its applicability to flight data from NASA’s dynamically-scaled generic transport model (GTM) aircraft.

Loss of control↗

NASA NTRSDate not supplied

High-Dimensional Similarity Search with Quantum-Assisted Variational Autoencoder

Recent progress in quantum algorithms and hardware indicates the potential importance of quantum computing in the near future. However, finding suitable application areas remains an active area of research. Quantum machine learning is touted as a potential approach to demonstrate quantum advantage within both the gate-model and the adiabatic schemes. For instance, the QVAE has been proposed as a quantum enhancement to the discrete VAE. We extend on previous work and study the real-world applicability of a QVAE by presenting a proof-of-concept for similarity search in large-scale high-dimensional datasets. While exact and fast similarity search algorithms are available for low dimensional datasets, scaling to high-dimensional data is non-trivial. We show how to construct a space-efficient search index based on the latent space representation of a QVAE. Our experiments show a correlation between the Hamming distance in the embedded space and the Euclidean distance in the original space on the MODIS dataset. Further, we find real-world speedups compared to linear search and demonstrate memory-efficient scaling to half a billion data points.

Data mining, similarity search, quantum machine le↗

NASA NTRS2011

Latent Virus Reactivation in Space Shuttle Astronauts

Latent virus reactivation was measured in 17 astronauts (16 male and 1 female) before, during, and after short-duration Space Shuttle missions. Blood, urine, and saliva samples were collected 2-4 months before launch, 10 days before launch (L-10), 2-3 hours after landing (R+0), 3 days after landing (R+14), and 120 days after landing (R+120). Epstein-Barr virus (EBV) DNA was measured in these samples by quantitative polymerase chain reaction. Varicella-zoster virus (VZV) DNA was measured in the 381 saliva samples and cytomegalovirus (CMV) DNA in the 66 urine samples collected from these subjects. Fourteen astronauts shed EBV DNA in 21% of their saliva samples before, during, and after flight, and 7 astronauts shed VZV in 7.4% of their samples during and after flight. It was interesting that shedding of both EBV and VZV increased during the flight phase relative to before or after flight. In the case of CMV, 32% of urine samples from 8 subjects contained DNA of this virus. In normal healthy control subjects, EBV shedding was found in 3% and VZV and CMV were found in less than 1% of the samples. The circadian rhythm of salivary cortisol measured before, during, and after space flight did not show any significant difference between flight phases. These data show that increased reactivation of latent herpes viruses may be associated with decreased immune system function, which has been reported in earlier studies as well as in these same subjects (data not reported here).

Mehta, S. K.↗

NASA NTRS2020

A Path Towards Quantum Advantage in Training Deep Generative Models with Quantum Annealing

A class of quantum-classical hybrid machine-learning algorithms can be obtained by integrating classical deep generative models with quantum probability distributions as 'priors' over their latent variables. We introduce a hybrid implementation of variational autoencoders (QVAE) and also present a technique to hybridize flow-based invertible generative models. We demonstrate the use of D-Wave quantum annealers as physical simulators of quantum Boltzmann machines (QBM) to perform quantum-assisted training of QVAE. Latent-space QBM develop slowly mixing modes, opening a path to obtain quantum advantage in generative modeling with available quantum devices.

Vinci, Walter↗

NASA NTRSDate not supplied

High dimensional similarity search with quantum assisted variational autoencoder

Recent progress in quantum algorithms and hardware is indicator of the potential importance of quantum computing in the next future. However, finding suitable application areas remains an active area of research. Quantum machine learning [1] is touted as a potential approach to demonstrate quantum advantage within both the gate-model [2,3] and the adiabatic [4,5] schemes. For instance, the Quantum-assisted Variational Autoencoder (QVAE) [6] has been proposed as a quantum enhancement to the discrete VAE [7]. We extend on previous work and study the real-world applicability of a QVAE, specifically, for similarity search in large-scale high dimensional datasets. While similarity search algorithms are available for low dimensional datasets, scaling to billion-scale datasets with thousands of dimensions is non-trivial. We show how the latent-space representation of a QVAE can be used to construct a space-efficient search index. We back up our claims by experimental results which show a correlation between the Hamming distance in the embedded space and the Euclidean distance in the original space on the Moderate Resolution Imaging Spectroradiometer (MODIS) dataset. Further, we show real-world speedups compared to linear search and demonstrate memory efficient scaling to large-scale datasets.

Nicholas D Gao↗

NASA NTRS2021

Loss of Control Detection for Commercial Transports Using Conditional Variational Autoencoders

This work describes a detector for the loss of control of a commercial transport in flight. The detector has a belief state defined by the latent variable stochastic modeling of a conditional variational autoencoder (CVAE) constructed with bidirectional recurrent layers. In 2000, the Boeing Company and the NASA Langley Research Center jointly developed a quantitative set of metrics for defining loss-of-control (LOC) for a commercial transport. We use the thresholds for these quantitative metrics to define a condition vector for training the CVAE. We demonstrate through experimentation that reconstruction probability is an accurate indicator that the vehicle has shifted to an LOC state. Second, we introduce a technique for inferring that the vehicle is experiencing a flight state change is approaching by measuring a shift in the sampling Gaussian distributions of the latent space. We provide an analysis of its applicability to flight data from a NASA generic commercial transport-type aircraft.

Newton H Campbell↗

NASA NTRSDate not supplied

Use of Design of Experiments in Determining Neural Network Architectures for Loss of Control Detection

We describe empirical methods for selecting a neural network architecture to implement belief state inference on generic commercial transport aircraft. We highlight a case study on the planning, execution, and analysis of a set of experiments to determine the configurations of a conditional variational autoencoder (CVAE). Our main contribution is the application of a structured method that can be used for machine learning in many aerospace applications. This method optimizes the structure and training parameters of a neural network for belief state inference, using Design of Experiments (DOE) statistical methodologies. The motivation for this specific DOE analysis was to identify the appropriate hyperparameters for measuring the CVAE reconstruction probability and latent space, such that the measurements can be used to infer qualitative state changes for the aircraft. We demonstrate that this process yields information about a trained neural network’s utility for this specific application, along with a quantifiable range of certainty. We execute 84 experiments using loss-of-control flight maneuver data from the NASA T-2 aircraft, demonstrating that this empirical process allows us to construct cheap and simple models with specific attributes amenable to belief state inference in aerospace applications.

Loss of Control↗

NASA NTRSDate not supplied

Use of Design of Experiments in Determining Neural Network Architectures for Loss of Control Detection

We describe empirical methods for selecting a neural network architecture to implement belief state inference on generic commercial transport aircraft. We highlight a case study on the planning, execution, and analysis of a set of experiments to determine the configurations of a conditional variational autoencoder (CVAE). Our main contribution is the application of a structured method that can be used for machine learning in many aerospace applications. This method optimizes the structure and training parameters of a neural network for belief state inference, using Design of Experiments (DOE) statistical methodologies. The motivation for this specific DOE analysis was to identify the appropriate hyperparameters for measuring the CVAE reconstruction probability and latent space, such that the measurements can be used to infer qualitative state changes for the aircraft. We demonstrate that this process yields information about a trained neural network’s utility for this specific application, along with a quantifiable range of certainty. We execute 84 experiments using loss-of-control flight maneuver data from the NASA T 2 aircraft, demonstrating that this empirical process allows us to construct cheap and simple models with specific attributes amenable to belief state inference in aerospace applications.

Loss of Control↗

NASA NTRSDate not supplied

Use of Design of Experiments in Determining Neural Network Architectures for Loss of Control Detection

Abstract—We describe empirical methods for selecting a neural network architecture to implement belief state inference on generic commercial transport aircraft. We highlight a case study on the planning, execution, and analysis of a set of experiments to determine the configurations of a conditional variational autoencoder (CVAE). Our main contribution is the application of a structured method that can be used for machine learning in many aerospace applications. This method optimizes the structure and training parameters of a neural network for belief state inference, using Design of Experiments (DOE) statistical methodologies. The motivation for this specific DOE analysis was to identify the appropriate hyperparameters for measuring the CVAE reconstruction probability and latent space, such that the measurements can be used to infer qualitative state changes for the aircraft. We demonstrate that this process yields information about a trained neural network’s utility for this specific application, along with a quantifiable range of certainty. We execute 84 experiments using loss-of-control flight maneuver data from the NASA T-2 aircraft, demonstrating that this empirical process allows us to construct cheap and simple models with specific attributes amenable to belief state inference in aerospace applications.

neural networks↗

NASA NTRS1999

Latent Viruses: A Space Travel Hazard??

A major issue associated with long-duration space flight is the possibility of infectious disease causing an unacceptable medical risk to crew members. Our proposal is designed to gain information that addresses several issues outlined in the Immunology/Infectious disease critical path. The major hypothesis addressed is that space flight causes alterations in the immune system that may allow latent viruses which are endogenous in the human population to reactivate and shed to higher levels than normal which can affect the health of crew members during a long term space-flight mission. We will initially focus our studies on the human herpesviruses and human polyomaviruses which are important pathogens known to establish latent infections in the human population. Both primary infection and reactivation from latent infection with this group of viruses can cause a variety of illnesses that result in morbidity and occasionally mortality of infected individuals. Effective vaccines exist for only one of the eight known human herpesviruses and the vaccine itself can still reactivate from latent infection. Available antivirals are of limited use and are effective against only a few of the human herpesviruses. Although most individuals display little if any clinical consequences from latent infection, events which alter immune function such as immunosuppressive therapy following solid organ transplantation are known to increase the risk of developing complications as a result of latent virus reactivation. This proposal will measure both the frequency and magnitude of viral shedding and genome loads in the blood from humans participating in activities that serve as ground based models of space flight conditions. Our initial goal is to develop sensitive quantitative competitive PCR- based assays (QC-PCR) to detect the herpesvirus Epstein-Barr virus (EBV), and the polyomaviruses SV40, BKV, and JCV. Using these assays we will establish baseline patterns of viral genome load in the blood and viral shedding from normal volunteers in a longitudinal study over I year in length. As a comparison, we will measure patterns of viral genome loads and shedding from individuals who are severely immunosuppressed, in whom herpesvirus reactivation or primary infection with a herpesvirus is known to cause complications. In addition, we will proceed to testing ground based analogs in collaboration with Dr. Duane Pierson (Lyndon B. Johnson Space Center). This will include measuring samples obtained from individuals living and working in the extreme environment of Antarctica. We expect to detect viral shedding or reactivation from most of the test groups, although the magnitude of shedding or reactivation cannot be predicted. The data accumulated from studies in this proposal should allow us to evaluate whether events that simulate certain aspects of space flight reactivate viral infections severe enough in nature that they may compromise the success of long-term space flight missions. These studies will also provide a foundation to monitor viral reactivation and shedding from crew members participating in actual space flight missions. We will present data showing the establishment of our QC-PCR assay for detection of EBV.

Ling, P. D.↗

NASA NTRS2010

Latent Virus Reactivation: From Space to Earth

Reactivation of latent viruses is a recognized consequence of decreased immunity. More recently viral reactivation has been identified as an important in vivo indicator of clinically relevant immune changes. Viral reactivation can be determined quickly and easily by the presence of virus in saliva and other body fluids. Real-time polymerase chain reaction (PCR) is a highly sensitive and specific molecular method to detect the presence of specific viral DNA. Studies in astronauts demonstrated that herpes simplex virus type 1(HSV-1), Epstein-Barr Virus (EBV), cytomegalovirus (CMV), and varicella zoster virus (VZV) reactivate at rates above normal during and after spaceflight in response to moderately decreased T-cell immunity. This technology was expanded to patients on Earth beginning with human immune deficiency virus (HIV) immuno-compromised patients. The HIV patients shed EBV in saliva at rates 9-fold higher than observed in astronauts demonstrating that the level of EBV shedding reflects the severity of impaired immunity. Whereas EBV reactivation is not expected to produce serious effects in astronauts on missions of 6 months or less, VZV reactivation in astronauts could produce shingles. Reactivation of live, infectious VZV in astronauts with no symptoms was demonstrated in astronauts during and after spaceflight. We applied our technology to study VZV-induced shingles in patients. In a study of 54 shingles patients, we showed salivary VZV was present in every patient on the day antiviral (acyclovir) treatment was initiated. Pain and skin lesions decreased with antiviral treatment. Corresponding decreases in levels of VZV were also observed and accompanied recovery. Although the level of VZV in shingles patients before the treatment was generally higher than those found in astronauts, lower range of VZV numbers in shingles patients overlapped with astronaut s levels. This suggests a potential risk of shingles to astronauts resulting from reactivation of VZV. In another clinical study of 25 shingles patients, PCR technology detected VZV in the serum and peripheral blood mononuclear cells of all 25 patients demonstrating for the first time that viremia is a common manifestation of herpes shingles.

Mehta, Satish K.↗

NASA NTRSDate not supplied

Exploring the Landscape of Earth and Space Science Informatics using Latent Topic Modeling

AGU Earth and Space Science Informatics (ESSI) is at the forefront of data management, analysis, large scale experimentation, and infrastructure development pertaining to Earth and Space Science interests. The key topics of interest within ESSI are also evolving and diversifying over time. We aim to observe and quantify the various topics covered in ESSI, analyze their trends over time, and identify the contributors’ affiliations to gain an understanding of the landscape of ESSI and the direction of the research and management. The data for this work are abstracts submitted to AGU’s ESSI Fall meeting; They serve as a proxy for key research and development areas within ESSI. We use an unsupervised topic modeling technique called Latent Dirichlet Allocation to observe the underlying topics covered in ESSI and their trends over time. With this presentation, we showcase our results from the analysis and insights gained.

Muthukumaran Ramasubramanian↗

NASA NTRS2002

Maintaining the Body's Immune System: Incidence of Latent Virus Shedding During Space Flight

Your body protects you from illness with its own security system - the immune system. This system keeps illness at bay not only by mounting a defense against foreign organisms, but also by controlling the population of bacteria and viruses that normally live in your body. But there's no need to panic: certain microbes can actually exist in your body without causing illness. Some bacteria are even beneficial - like the E. coli in the large intestine that are an important source of vitamin K. While viruses are not exactly considered beneficial, they can also inhabit the human body without causing immediate harm or infection. A good example is the herpes simplex virus type 1 (HSV1), more commonly known as cold sores or fever blisters. This virus infects 70 to 80 percent of all adults but remains latent much of the time. While latent, the virus within cells remains dormant. Activation of the dormant virus causes it to make copies of itself (known as replication) constantly detectable in body fluids such as urine or saliva in a process called shedding. When a person becomes sick or stressed, however, this weakened condition allows the virus to reactivate and multiply. These elevated levels may be enough to produce symptoms, but shedding can also occur without symptoms. This ability to shed without showing signs of infection, or asymptomatic shedding, is of great interest, as it increases the chances of infecting others. The stresses associated with space flight - adapting to microgravity, isolation from family and friends, living and working in a confined space, sleep deprivation, and busy schedules, to name but a few - may weaken astronauts' immune systems, leaving them at greater risk of viral reactivation. Members of the STS-107 crew will participate in this experiment, Incidence of Latent Viral Shedding in Space Flight, to help scientists understand how reactivation works in space, and at what level replication reaches before symptoms begin to show. This study also promises more insight into the behavior of the larger virus family, herpesvirus, which will help us understand how to prevent infection in populations on Earth and reactivation in those already infected.

Pierson, Duane↗

NASA NTRS2008

Incidence of Latent Virus Shedding during Space Flight

Measurements of immune parameters of both cellular and innate immunity indicate alterations in immune function in astronauts. Immune changes are due to stress and perhaps other factors associated with launch, flight, and landing phases. Medical relevance of observed changes is not known. The reactivation of latent viruses has been identified as an important in vivo indicator of clinically relevant immune changes. The polymerase chain reaction (PCR) was used to detect the presence of specific viral DNA in body fluids. Initial studies demonstrated Epstein-Barr virus (EBV) reactivation during all 3 mission phases. EBV is shed in saliva following reactivation from B-cells. Incidence of EBV in saliva was higher than control subjects during all 3 mission phases. However, quantitative PCR revealed 10-fold higher levels of EBV DNA present in saliva collected during flight than found in pre- and post flight specimens. To determine if other latent viruses showed similar effects, cytomegalovirus (CMV), another herpes virus, shed in urine following reactivation was studied. A very low incidence (less than 2%) of CMV in urine is found in healthy, lowstressed individuals. However, 25-50% of astronauts shed CMV in their urine before, during, or after flight. Our studies are now focused on varicella-zoster virus (VZV), the etiological agent of chicken-pox during childhood and shingles later in life. We demonstrated reactivation of VZV and shedding of the virus during and after spaceflight in saliva of astronauts with no sign of active infection or symptoms. The maximum shedding of VZV occurred during the flight phase and diminishes rapidly during the first five days after landing. We have utilized the same PCR assay for VZV in a clinical study of shingles patients. Generally, shingles patients shed much more VZV in saliva than astronauts. However, the VZV levels in astronauts overlap with the lower range of VZV numbers in shingles patients. Saliva from shingles patients and astronauts were cultured and infectious VZV was recovered from both groups. We have concluded that multiple latent viruses do reactivate before, during, and after spaceflight and serve as very sensitive indicators for diminished cellular immunity. Future plans will be focused on the clinical risks posed by the reactivation of these viruses. Initial efforts will determine the effect of longer missions on the International Space Station on the reactivation patterns of these viruses.

Mehta, Satish K.↗

NASA NTRS1999

Lytic Replication of Epstein-Barr Virus During Space Flight

Reactivation of latent Epstein-Barr virus (EBV) may be an important threat to crew health during extended space missions. Cellular immunity, which is decreased during and after space flight, is responsible for controlling EBV replication in vivo. In this study, we investigated the effects of short-term space flight on latent EBV reactivation.

Stowe, R. P.↗

NASA NTRS2001

Space Flight-Induced Reactivation of Latent Epstein-Barr Virus

Reactivation of latent Epstein-Barr virus (EBV) may be an important threat to crew health during extended space missions. Decreased cellular immune function has been reported both during and after space flight. Preliminary studies have demonstrated increased EBV shedding in saliva as well as increased antibody titers to EBV lytic proteins. We hypothesize that the combined effects of microgravity along with associated physical and psychological stress will decrease EBV-specific T-cell immunity and reactivate latent EBV in infected B-lymphocytes. If increased virus production and clonal expansion of infected B-lymphocytes are detected, then pharmacological measures can be developed and instituted prior to onset of overt clinical disease. More importantly, we will begin to understand the basic mechanisms involved in stress-induced reactivation of EBV in circulating B-lymphocytes.

Stowe, Raymond P.↗