skip to main content
OSTI.GOV title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: Selecting an Informative/Discriminating Multivariate Response for Inverse Prediction

Abstract

The inverse prediction is important in a variety of scientific and engineering applications, such as to predict properties/characteristics of an object by using multiple measurements obtained from it. Inverse prediction can be accomplished by inverting parameterized forward models that relate the measurements (responses) to the properties/characteristics of interest. Sometimes forward models are computational/science based; but often, forward models are empirically based response surface models, obtained by using the results of controlled experimentation. For empirical models, it is important that the experiments provide a sound basis to develop accurate forward models in terms of the properties/characteristics (factors). And while nature dictates the causal relationships between factors and responses, experimenters can control the complexity, accuracy, and precision of forward models constructed via selection of factors, factor levels, and the set of trials that are performed. Recognition of the uncertainty in the estimated forward models leads to an errors-in-variables approach for inverse prediction. The forward models (estimated by experiments or science based) can also be used to analyze how well candidate responses complement one another for inverse prediction over the range of the factor space of interest. Furthermore, one may find that some responses are complementary, redundant, or noninformative. Simple analysis and examplesmore » illustrate how an informative and discriminating subset of responses could be selected among candidates in cases where the number of responses that can be acquired during inverse prediction is limited by difficulty, expense, and/or availability of material.« less

Authors:
 [1];  [1]; ORCiD logo [2];  [2]; ORCiD logo [2]
  1. Sandia National Lab. (SNL-NM), Albuquerque, NM (United States)
  2. Los Alamos National Lab. (LANL), Los Alamos, NM (United States)
Publication Date:
Research Org.:
Los Alamos National Lab. (LANL), Los Alamos, NM (United States)
Sponsoring Org.:
USDOE
OSTI Identifier:
1375866
Report Number(s):
LA-UR-15-27620
Journal ID: ISSN 0022-4065
Grant/Contract Number:
AC52-06NA25396
Resource Type:
Journal Article: Accepted Manuscript
Journal Name:
Journal of Quality Technology
Additional Journal Information:
Journal Volume: 49; Journal Issue: 3; Journal ID: ISSN 0022-4065
Publisher:
American Society for Quality (ASQ)
Country of Publication:
United States
Language:
English
Subject:
97 MATHEMATICS AND COMPUTING; Prediction; Multivariate analysis; Measurement error; Regression analysis; Least squares

Citation Formats

Thomas, Edward V., Lewis, John. R., Anderson-Cook, Christine Michaela, Hamada, Michael Scott, and Burr, Thomas Lee. Selecting an Informative/Discriminating Multivariate Response for Inverse Prediction. United States: N. p., 2017. Web.
Thomas, Edward V., Lewis, John. R., Anderson-Cook, Christine Michaela, Hamada, Michael Scott, & Burr, Thomas Lee. Selecting an Informative/Discriminating Multivariate Response for Inverse Prediction. United States.
Thomas, Edward V., Lewis, John. R., Anderson-Cook, Christine Michaela, Hamada, Michael Scott, and Burr, Thomas Lee. 2017. "Selecting an Informative/Discriminating Multivariate Response for Inverse Prediction". United States. doi:.
@article{osti_1375866,
title = {Selecting an Informative/Discriminating Multivariate Response for Inverse Prediction},
author = {Thomas, Edward V. and Lewis, John. R. and Anderson-Cook, Christine Michaela and Hamada, Michael Scott and Burr, Thomas Lee},
abstractNote = {The inverse prediction is important in a variety of scientific and engineering applications, such as to predict properties/characteristics of an object by using multiple measurements obtained from it. Inverse prediction can be accomplished by inverting parameterized forward models that relate the measurements (responses) to the properties/characteristics of interest. Sometimes forward models are computational/science based; but often, forward models are empirically based response surface models, obtained by using the results of controlled experimentation. For empirical models, it is important that the experiments provide a sound basis to develop accurate forward models in terms of the properties/characteristics (factors). And while nature dictates the causal relationships between factors and responses, experimenters can control the complexity, accuracy, and precision of forward models constructed via selection of factors, factor levels, and the set of trials that are performed. Recognition of the uncertainty in the estimated forward models leads to an errors-in-variables approach for inverse prediction. The forward models (estimated by experiments or science based) can also be used to analyze how well candidate responses complement one another for inverse prediction over the range of the factor space of interest. Furthermore, one may find that some responses are complementary, redundant, or noninformative. Simple analysis and examples illustrate how an informative and discriminating subset of responses could be selected among candidates in cases where the number of responses that can be acquired during inverse prediction is limited by difficulty, expense, and/or availability of material.},
doi = {},
journal = {Journal of Quality Technology},
number = 3,
volume = 49,
place = {United States},
year = 2017,
month = 7
}

Journal Article:
Free Publicly Available Full Text
This content will become publicly available on July 1, 2018
Publisher's Version of Record
The DOI is not currently available

Save / Share:
  • Abstract not provided.
  • Purpose: Vascular injury could be a cause of hippocampal dysfunction leading to late neurocognitive decline in patients receiving brain radiotherapy (RT). Hence, our aim was to develop a multivariate interaction model for characterization of hippocampal vascular dose-response and early prediction of radiation-induced late neurocognitive impairments. Methods: 27 patients (17 males and 10 females, age 31–80 years) were enrolled in an IRB-approved prospective longitudinal study. All patients were diagnosed with a low-grade glioma or benign tumor and treated by 3-D conformal or intensity-modulated RT with a median dose of 54 Gy (50.4–59.4 Gy in 1.8− Gy fractions). Six DCE-MRI scans weremore » performed from pre-RT to 18 months post-RT. DCE data were fitted to the modified Toft model to obtain the transfer constant of gadolinium influx from the intravascular space into the extravascular extracellular space, Ktrans, and the fraction of blood plasma volume, Vp. The hippocampus vascular property alterations after starting RT were characterized by changes in the hippocampal mean values of, μh(Ktrans)τ and μh(Vp)τ. The dose-response, Δμh(Ktrans/Vp)pre->τ, was modeled using a multivariate linear regression considering integrations of doses with age, sex, hippocampal laterality and presence of tumor/edema near a hippocampus. Finally, the early vascular dose-response in hippocampus was correlated with neurocognitive decline 6 and 18 months post-RT. Results: The μh(Ktrans) increased significantly from pre-RT to 1 month post-RT (p<0.0004). The multivariate model showed that the dose effect on Δμh(Ktrans)pre->1M post-RT was interacted with sex (p<0.0007) and age (p<0.00004), with the dose-response more pronounced in older females. Also, the vascular dose-response in the left hippocampus of females was significantly correlated with memory function decline at 6 (r = − 0.95, p<0.0006) and 18 (r = −0.88, p<0.02) months post-RT. Conclusion: The hippocampal vascular response to radiation could be sex and age dependent. The early hippocampal vascular dose-response could predict late neurocognitive dysfunction. (Support: NIH-RO1NS064973)« less
  • A number of interesting problems in the design of experiments such as sensor allocation, selection of sites for the observing stations, determining sampler positions in traffic monitoring, and which variables to survey/measure in sampling studies may be considered in the following setting: Given a covariance matrix of multi-dimension random vector and given a ratio of the number of possible observations to the observational error select those components which must be observed to guarantee minimization of an objective function describing the quality of prediction of all or prescribed components. The authors show that the problem can be considered in the frameworkmore » of convex design theory and derive some simple but effective algorithm for selection of an optimal subset of components to be observed.« less
  • The excesses of the cosmic positron fraction recently measured by PAMELA and the electron spectra by ATIC, PPB-BETS, Fermi, and H.E.S.S. indicate the existence of primary electron and positron sources. The possible explanations include dark matter annihilation, decay, and astrophysical origin, like pulsars. In this work we show that these three scenarios can all explain the experimental results of the cosmic e{sup {+-}} excess. However, it may be difficult to discriminate these different scenarios by the local measurements of electrons and positrons. We propose possible discriminations among these scenarios through the synchrotron and inverse Compton radiation of the primary electrons/positronsmore » from the region close to the Galactic center. Taking typical configurations, we find the three scenarios predict quite different spectra and skymaps of the synchrotron and inverse Compton radiation, though there are relatively large uncertainties. The most prominent differences come from the energy band 10{sup 4}{approx}10{sup 9} MHz for synchrotron emission and > or approx. 10 GeV for inverse Compton emission. It might be able to discriminate at least the annihilating dark matter scenario from the other two given the high precision synchrotron and diffuse {gamma}-ray skymaps in the future.« less
  • Reliable fire detection is essential for both safe evacuation and containment or extinguishment. In order to increase reliability by reducing the number of nuisance fire alarms in underground mines that use diesel-powered equipment, the U.S. Bureau of Mines has developed a diesel-discriminating fire detector (DDD). It was designed to discriminate between smoke produced by a fire and the smoke-laden exhaust of a diesel engine. Experiments were conducted by the Bureau to compare the smoke detection capabilities of the DDD with those of conventional fire detectors in response to smoldering coal and conveyor belting. A comparison was made among the alarmmore » times of a carbon monoxide (CO) detector with an alarm threshold of 5 ppm, a smoke detector with an optical density alarm threshold of 0.044 m[sup [minus]1], and the DDD with an alarm threshold of 0.025 V. The results show that the DDD will reliably detect developing coal and conveyor belt fires. The average time delay separating the DDD alarm from the first detector to alarm was 76 s for smoldering conveyor belt and 65 s for smoldering coal. The longest time delay between the response of the DDD and the first detector to alarm was approximately 120 s.« less