Search NASA⌕ Search

SEARCH · Search NASA

Results for “Poisson regression”

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 19 records

Zero-truncated Poisson regression for sparse multiway count data corrupted by false zeros

Abstract We propose a novel statistical inference methodology for multiway count data that is corrupted by false zeros that are indistinguishable from true zero counts. Our approach consists of zero-truncating the Poisson distribution to neglect all zero values. This simple truncated approach dispenses with the need to distinguish between true and false zero counts and reduces the amount of data to be processed. Inference is accomplished via tensor completion that imposes low-rank tensor structure on the Poisson parameter space. Our main result shows that an $N$-way rank-$R$ parametric tensor $\boldsymbol{\mathscr{M}}\in (0,\infty )^{I\times \cdots \times I}$ generating Poisson observations can be accurately estimated by zero-truncated Poisson regression from approximately $IR^2\log _2^2(I)$ non-zero counts under the nonnegative canonical polyadic decomposition. Our result also quantifies the error made by zero-truncating the Poisson distribution when the parameter is uniformly bounded from below. Therefore, under a low-rank multiparameter model, we propose an implementable approach guaranteed to achieve accurate regression in under-determined scenarios with substantial corruption by false zeros. Several numerical experiments are presented to explore the theoretical results.

97 MATHEMATICS AND COMPUTING↗

Poisson Regression Analysis of Illness and Injury Surveillance Data

The Department of Energy (DOE) uses illness and injury surveillance to monitor morbidity and assess the overall health of the work force. Data collected from each participating site include health events and a roster file with demographic information. The source data files are maintained in a relational data base, and are used to obtain stratified tables of health event counts and person time at risk that serve as the starting point for Poisson regression analysis. The explanatory variables that define these tables are age, gender, occupational group, and time.

60 APPLIED LIFE SCIENCES↗

Evaluating county-level lung cancer incidence from environmental radiation exposure, PM 2.5 , and other exposures with regression and machine learning models

Characterizing the interplay between exposures shaping the human exposome is vital for uncovering the etiology of complex diseases. For example, cancer risk is modified by a range of multifactorial external environmental exposures. Environmental, socioeconomic, and lifestyle factors all shape lung cancer risk. However, epidemiological studies of radon aimed at identifying populations at high risk for lung cancer often fail to consider multiple exposures simultaneously. For example, moderating factors, such as PM 2.5 , may affect the transport of radon progeny to lung tissue. This ecological analysis leveraged a population-level dataset from the National Cancer Institute’s Surveillance, Epidemiology, and End-Results data (2013–17) to simultaneously investigate the effect of multiple sources of low-dose radiation (gross γ activity and indoor radon) and PM 2.5 on lung cancer incidence rates in the USA. County-level factors (environmental, sociodemographic, lifestyle) were controlled for, and Poisson regression and random forest models were used to assess the association between radon exposure and lung and bronchus cancer incidence rates. Tree-based machine learning (ML) method perform better than traditional regression: Poisson regression: 6.29/7.13 (mean absolute percentage error, MAPE), 12.70/12.77 (root mean square error, RMSE); Poisson random forest regression: 1.22/1.16 (MAPE), 8.01/8.15 (RMSE). The effect of PM 2.5 increased with the concentration of environmental radon, thereby confirming findings from previous studies that investigated the possible synergistic effect of radon and PM 2.5 on health outcomes. In summary, the results demonstrated (1) a need to consider multiple environmental exposures when assessing radon exposure’s association with lung cancer risk, thereby highlighting (1) the importance of an exposomics framework and (2) that employing ML models may capture the complex interplay between environmental exposures and health, as in the case of indoor radon exposure and lung cancer incidence.

63 RADIATION, THERMAL, AND OTHER ENVIRON. POLLUTAN↗

Bayesian probability of predictive agreement for comparing the outcome of two separate regressions

The concept of a Bayesian probability of agreement was recently introduced to give the posterior probabilities that the response surfaces for two different groups are within δ of one another. For example, a difference of less than δ in the mean response at fixed levels of the predictor variables might be thought to be practically unimportant. In such a case, we would say that the mean responses are in agreement. The posterior probability of this is called the Bayesian probability of agreement. In this article, we quantify the probability that new response observations from two groups will be within δ for a continuous response, and the probability that the two responses agree completely for categorical cases such as logistic regression and Poisson regression. We call these Bayesian comparative predictive probabilities, with the former being the predictive probability of agreement. We use Markov chain Monte Carlo simulation to estimate the posterior distribution of the model parameters and then the predictive probability of agreement. Here, we illustrate the use of this methodology with three examples and provide a freely available R Shiny app that automates the computation and estimation associated with the methodology.

97 MATHEMATICS AND COMPUTING↗

Prescribed fires, smoke exposure, and hospital utilization among heart failure patients

Abstract Background Prescribed fires often have ecological benefits, but their environmental health risks have been infrequently studied. We investigated associations between residing near a prescribed fire, wildfire smoke exposure, and heart failure (HF) patients’ hospital utilization. Methods We used electronic health records from January 2014 to December 2016 in a North Carolina hospital-based cohort to determine HF diagnoses, primary residence, and hospital utilization. Using a cross-sectional study design, we associated the prescribed fire occurrences within 1, 2, and 5 km of the patients’ primary residence with the number of hospital visits and 7- and 30-day readmissions. To compare prescribed fire associations with those observed for wildfire smoke, we also associated zip code-level smoke density data designed to capture wildfire smoke emissions with hospital utilization amongst HF patients. Quasi-Poisson regression models were used for the number of hospital visits, while zero-inflated Poisson regression models were used for readmissions. All models were adjusted for age, sex, race, and neighborhood socioeconomic status and included an offset for follow-up time. The results are the percent change and the 95% confidence interval (CI). Results Associations between prescribed fire occurrences and hospital visits were generally null, with the few associations observed being with prescribed fires within 5 and 2 km of the primary residence in the negative direction but not the more restrictive 1 km radius. However, exposure to medium or heavy smoke (primarily from wildfires) at the zip code level was associated with both 7-day (8.5% increase; 95% CI = 1.5%, 16.0%) and 30-day readmissions (5.4%; 95% CI = 2.3%, 8.5%), and to a lesser degree, hospital visits (1.5%; 95% CI: 0.0%, 3.0%) matching previous studies. Conclusions Area-level smoke exposure driven by wildfires is positively associated with hospital utilization but not proximity to prescribed fires.

Raab, Henry↗

Pollen-mediated gene flow from transgenic to non-transgenic switchgrass (Panicum virgatum L.) in the field

Background: Switchgrass is C4 perennial grass species that is being developed as a cellulosic bioenergy feedstock. It is wind-pollinated and considered to be an obligate outcrosser. Genetic engineering has been used to alter cell walls for more facile bioprocessing and biofuel yield. Gene flow from transgenic cultivars would likely be of regulatory concern. In this study we investigated pollen-mediated gene flow from transgenic to nontransgenic switchgrass in a 3-year field experiment performed in Oliver Springs, Tennessee, U.S.A. using a modified Nelder wheel design. The planted area (0.6 ha) contained sexually compatible pollen source and pollen receptor switchgrass plants. One hundred clonal switchgrass ‘Alamo’ plants transgenic for an orange-fluorescent protein (OFP) and hygromycin resistance were used as the pollen source; whole plants, including pollen, were orange-fluorescent. To assess pollen movement, pollen traps were placed at 10 m intervals from the pollen-source plot in the four cardinal directions extending to 20 m, 30 m, 30 m, and 100 m to the north, south, west, and east, respectively. To assess pollination rates, nontransgenic ‘Alamo 2’ switchgrass clones were planted in pairs adjacent to pollen traps. Results: In the eastward direction there was a 98% decrease in OFP pollen grains from 10 to 100 m from the pollen-source plot (Poisson regression, F1,8 = 288.38, P < 0.0001). At the end of the second and third year, 1,820 F1 seeds were collected from pollen recipient-plots of which 962 (52.9%) germinated and analyzed for their transgenic status. Transgenic progeny production detected in each pollen-recipient plot decreased with increased distance from the edge of the transgenic plot (Poisson regression, F1,15 = 12.98, P < 0.003). The frequency of transgenic progeny detected in the eastward plots (the direction of the prevailing wind) ranged from 79.2% at 10 m to 9.3% at 100 m. Conclusions: In these experiments we found transgenic pollen movement and hybridization rates to be inversely associated with distance. However, these data suggest pollen-mediated gene flow is likely to occur up to, at least, 100 m. This study gives baseline data useful to determine isolation distances and other management practices should transgenic switchgrass be grown commercially in relevant environments.

09 BIOMASS FUELS↗

Zero-Truncated Poisson Tensor Decomposition for Sparse Count Data

We propose a novel statistical inference paradigm for zero-inflated multiway count data that dispenses with the need to distinguish between true and false zero counts. Our approach ignores all zero entries and applies zero-truncated Poisson regression on the positive counts. Inference is accomplished via tensor completion that imposes low-rank structure on the Poisson parameter space. Our main result shows that an $\textit{N}$-way rank-R parametric tensor 𝓜 ϵ (0, ∞) $I$Χ∙∙∙Χ$I$ generating Poisson observations can be accurately estimated from approximately $IR^2 \text{log}^2_2(I)$ non-zero counts for a nonnegative canonical polyadic decomposition. Several numerical experiments are presented demonstrating that our zero-truncated paradigm is comparable to the ideal scenario where the locations of false zero counts are known $\textit{a priori}$.

97 MATHEMATICS AND COMPUTING↗

Bayesian probability of agreement for comparing the similarity of response surfaces

For a variety of linear and generalized linear models, there may be interest in evaluating whether the expected response values in two different populations are sufficiently similar to be considered practically equivalent. The probability of agreement has been introduced as a strategy for quantifying the similarity between two groups. We propose a Bayesian version of the probability of agreement to quantify the similarity between linear and generalized linear response surfaces. The proposed methodology is based on Markov chain Monte Carlo estimation and it allows for an intuitive interpretation based directly on population parameters. As demonstrated, the methodology can be flexibly applied to a variety of different models. We illustrate its use with three examples for which the response depends on predictor variables through a linear, logistic, or Poisson regression model. The computation associated with this approach has been automated with a freely available R Shiny app.

97 MATHEMATICS AND COMPUTING↗

A Critical Examination of Figure of Merit (FOM). Assessing the Goodness-of-Fit in Gamma/X-ray Peak Analysis

In this paper we develop and investigate several criteria for assessing how well a proposed spectral form fits observed spectra. We consider the classical improved figure of merit (FOM) along with several modifications, as well as criteria motivated by Poisson regression from the statistical literature. We also develop a new FOM that is based on the statistical idea of the bootstrap. A spectral simulator has been developed to assess the performance of these different criteria under multiple data configurations.

97 MATHEMATICS AND COMPUTING↗

Tensor Decompositions for Count Data that Leverage Stochastic and Deterministic Optimization

There is growing interest to extend low-rank matrix decompositions to multi-way arrays, or tensors. One fundamental low-rank tensor decomposition is the canonical polyadic decomposition (CPD). The challenge of fitting a low-rank, nonnegative CPD model to Poisson-distributed count data is of particular interest. Several popular algorithms use local search methods to approximate the global maximum likelihood estimator from local minima. Simultaneously, a recent trend in theoretical computer science and numerical linear algebra leverages randomization to solve very large, hard problems. The typical approach is to use randomization for a fast approximation and determinism for refinement to yield effective algorithms with theoretical guarantees. Two popular algorithms for Poisson CPD reflect that emergent dichotomy: CP Alternating Poisson Regression is a deterministic algorithm and Generalized Canonical Polyadic decomposition makes use of stochastic algorithms in several variants. This work extends recent work to develop two new methods that leverage randomized and deterministic algorithms for improved accuracy and performance.

97 MATHEMATICS AND COMPUTING↗

Tensor decompositions for count data that leverage stochastic and deterministic optimization

There is growing interest to extend low-rank matrix decompositions to multi-way arrays, or tensors. One fundamental low-rank tensor decomposition is the canonical polyadic decomposition (CPD). The challenge of fitting a low-rank, nonnegative CPD model to Poisson-distributed count data is of particular interest. Several popular algorithms use local search methods to approximate the maximum likelihood estimator (MLE) of the Poisson CPD model. Here, this work presents two new algorithms that extend state-of-the-art local methods for Poisson CPD. Hybrid GCP-CPAPR combines Generalized Canonical Decomposition (GCP) with stochastic optimization and CP Alternating Poisson Regression (CPAPR), a deterministic algorithm, to increase the probability of converging to the MLE over either method used alone. Restarted CPAPR with SVDrop uses a heuristic based on the singular values of the CPD model unfoldings to identify convergence toward optimizers that are not the MLE and restarts within the feasible domain of the optimization problem, thus reducing overall computational cost when using a multi-start strategy. We provide empirical evidence that indicates our approaches outperform existing methods with respect to converging to the Poisson CPD MLE.

CPAPR↗

Correction of confidence intervals in excess relative risk models using Monte Carlo dosimetry systems with shared errors

Introduction: Measurement errors are ubiquitous in epidemiological studies, especially for doses arising from environmental and occupational exposures, which are difficult to assess. Correction methods for independent measurement errors, have been comprehensively described elsewhere, including regression calibration, simulation extrapolation (SIMEX), etc. However, when exposure estimates are constructed using complicated physical and biological models, uncertainties in the dosimetry system can become very complex. In such systems, uncertainties in some parameters may affect a large group of study participants simultaneously, and hence are “shared” as opposed to independent dosimetric uncertainties. Representing the structure of the uncertainty in the dosimetry by providing multiple realizations of possible dose has been described for several studies. The goal of this paper is to implement, and explore the properties of, a generalized estimating equation-based approach to shared uncertainties based on principles described in an earlier paper. Methods: The effects of shared and unshared dosimetry error on parameter estimates and confidence intervals in a generalized excess relative risk model are examined. A post-hoc correction method is implemented that constructs corrected confidence intervals based on an underlying variance structure described in an earlier paper. In this correction method it is assumed that the estimated slope parameter in a linear excess relative risk model is distributed as a mixture of normal and lognormal components. A simulation study of the behavior of this correction was developed based on multiple realizations provided by the Monte Carlo dosimetry system used for the plutonium exposures in the Mayak Worker Cohort and a risk model with close similarity to that used in analyses of lung cancer data in that cohort. Results: In the simulation study we found that shared dose errors severely degraded the coverage probabilities of confidence limits for the slope parameter in the linear excess relative risk model relating uncertain exposure to lung cancer risk. Using our correction method to adjust the confidence intervals provided much improved coverage probabilities for this parameter. A smaller amount of degradation in confidence interval coverage for other parameters (e.g. those in the baseline and dose modification terms) was noted, and the proposed correction method also improved coverage for these parameters as well. Discussion: Compared to other recently described methods our proposed correction method relies on relatively familiar software to fit dose response models. The basic modeling philosophy described here first uses existing Poisson regression methods to estimate parameters in the excess relative risk model; in this first stage mean dose (computed from the multiple realizations) is used as the independent variable in the regression. Once the model is fit we then correct (widen) the standard errors for all parameters using the proposed correction method. The simulation studies indicate that the methods described here are an important contribution to the available tools for dealing with shared measurement errors in complex risk analysis.

61 RADIATION PROTECTION AND DOSIMETRY↗

Associations between cumulative environmental quality and ten selected birth defects in Texas

Abstract Background Causes of most birth defects are largely unknown. Genetics, maternal factors (e.g., age, smoking) and environmental exposures have all been linked to some birth defects, including neural tube, oral cleft, limb reduction, and gastroschisis; however, the contribution of cumulative exposures across several environmental domains in association with these defects is not well understood. Methods The Environmental Quality Index (EQI) and its domains (air, water, land, sociodemographic, built) were used to estimate county‐level cumulative environmental exposures from 2006–2010 and matched to birth defects identified from Texas Birth Defects Registry and live birth records from births in years 2007–2010 ( N = 1,610,709). Poisson regression models estimated prevalence ratios (PR) and 95% confidence intervals (CI) for associations between 10 birth defects and the EQI. Results We observed some positive associations between worst environmental quality and neural tube, anencephaly, spina bifida, oral cleft, cleft palate, cleft lip with and without cleft palate, and gastroschisis [PR range: 1.12–1.55], but near null associations with limb reduction defects. Among domain specific results, we observed the strongest positive associations with the sociodemographic domain across birth defects but varied positive associations among the air and water domains, and negative or null associations with the land and built domains. Overall, few exposure‐response patterns were evident. Conclusions Our results highlight the complexities of cumulative, simultaneous environmental exposures in the prevalence rates of 10 selected birth defects. We were able to explore the impact of overall and domain specific environmental quality on birth defects and identify potential domain specific drivers of these associations.

Krajewski, Alison K.↗

Radiation risk of incident colorectal cancer by anatomical site among atomic bomb survivors: 1958–2009

Radiation effects on colorectal cancer rates, adjusted for smoking, alcohol intake and frequency of meat consumption and body mass index (BMI) by anatomical subsite (proximal colon, distal colon and rectum) were examined in a cohort of 105,444 atomic bomb survivors. Poisson regression methods were used to describe radiation‐associated excess relative risks (ERR) and excess absolute rates (EAR) for the 1958–2009 period. There were 2,960 first primary colorectal cancers including 894 proximal, 871 distal and 1,046 rectal cancers. Smoking, alcohol intake and BMI were associated with subsite‐specific cancer background rates. Significant linear dose–responses were found for total colon (sex‐averaged ERR/Gy for 70 years old exposed at age 30 = 0.63, 95% confidence interval [CI]: 0.34; 0.98), proximal [ERR = 0.80, 95% CI: 0.32; 1.44] and distal colon cancers [ERR = 0.50, 95% CI: 0.04; 0.97], but not for rectal cancer [ERR = 0.023, 95% CI: −0.081; 0.13]. The ERRs for proximal and distal colon cancers were not significantly different ( p = 0.41). The ERR decreased with attained age for total colon, but not for proximal colon cancer, and with calendar year for distal colon cancer. The ERRs and EARs did not vary by age at exposure, except for decreasing trend in EAR for proximal colon cancer. In conclusion, ionizing radiation is associated with increased risk of proximal and distal colon cancers. The ERR for proximal cancer persists over time, but that for distal colon cancer decreases. There continues to be no indication of radiation effects on rectal cancer incidence in this population.

60 APPLIED LIFE SCIENCES↗

Misclassification of primary liver cancer in the Life Span Study of atomic bomb survivors

Primary liver cancer is difficult to diagnose accurately at death, due to metastases from nearby organs and to concomitant diseases, such as chronic hepatitis and cirrhosis. Trends in diagnostic accuracy could affect radiation risk estimates for incident liver cancer by altering background rates or by impacting risk modification by sex and age. We quantified the potential impact of death‐certificate inaccuracies on radiation risk estimates for liver cancer in the Life Span Study of atomic bomb survivors. True‐positive and false‐negative rates were obtained from a previous study that compared death‐certificate causes of death with those based on pathological review, from 1958 to 1987. We assumed various scenarios for misclassification rates after 1987. We obtained estimated true positives and estimated false negatives by stratified sampling from binomial distributions with probabilities given by the true‐positive and false‐negative rates, respectively. Poisson regression methods were applied to highly stratified person‐year tables of corrected case counts and accrued person years. During the study period (1958–2009), there were 1,885 cases of liver cancer, which included 383 death‐certificate‐only (DCO) cases; 1,283 cases with chronic liver disease as the underlying cause of death; and 150 DCO cases of pancreatic cancer among 105,444 study participants. Across the range of scenarios considered, radiation risk estimates based on corrected case counts were attenuated, on average, by 13–30%. Our results indicated that radiation risk estimates for liver cancer were potentially sensitive to death‐certificate inaccuracies. Additional data are needed to inform misclassification rates in recent years.

French, Benjamin↗