SEARCH · Search NASA

Results for “quantile 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.

SourceResearch connectionFrom yearThrough year

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

83 records · Page 5

All sources NASA NTRS DOE OSTI

NASA NTRS1985

Tolerance bounds for log gamma regression models

The present procedure for finding lower confidence bounds for the quantiles of Weibull populations, on the basis of the solution of a quadratic equation, is more accurate than current Monte Carlo tables and extends to any location-scale family. It is shown that this method is accurate for all members of the log gamma(K) family, where K = 1/2 to infinity, and works well for censored data, while also extending to regression data. An even more accurate procedure involving an approximation to the Lawless (1982) conditional procedure, with numerical integrations whose tables are independent of the data, is also presented. These methods are applied to the case of failure strengths of ceramic specimens from each of three billets of Si3N4, which have undergone flexural strength testing.

Jones, R. A.↗

DOE OSTI2013

Adjusting inhomogeneous daily temperature variability using wavelet analysis

A wavelet-analysis-based homogenization method (WHM) is developed for detecting and adjusting biases of variability in a daily climate observation series, which are influential in the estimation of climate extremes and relevant trends in the time series. WHM is applied to the Central England daily mean temperature series (CET, 1772–2010) and the daily mean temperature series at Henan station in mountainous western China (HNT, 1960–2008) in order to demonstrate the usefulness of the new technique. The changes of methods for calculating daily mean temperature for CET in 1878 caused significantly reduced daily variability (DV) in the subsequent sub-series of CET. The relocation of the Chinese station in 1981 to a higher mountainous site resulted in not only enhanced DV but also in weather (weekly) time-scale variability (WV) in the subsequent sub-series of HNT. The adjustments based on WHM in the present cases improve the estimates of the long-term trends of climate extremes such as hot days, cold days, heat waves, and cold surges. A few widely applied methods of homogenization such as multiple analysis of series for homogenization (MASH), higher-order moment (HOM), RHtestsV3 with quantile-matching (QM) adjustments and two-phase regression (TPR) are also applied to HNT for comparison. TPR is not aimed at improving the homogeneity of variability in the time series; MASH does not improve daily variability either; HOM improves it in a small way; and with RHtestsV3 with QM adjustments it is considerably improved, but biases remain. In contrast, WHM improves the homogeneity of the short-term variability in the time series, resulting in reasonable assessments of climate extremes and trends in the daily observations.

54 ENVIRONMENTAL SCIENCES↗

DOE OSTI2025

Data and scripts associated with “Allometric scaling of hyporheic respiration across basins in the Pacific Northwest USA"

This data package is associated with the publication “Allometric scaling of hyporheic respiration across basins in the Pacific Northwest USA” submitted to JGR-Biogeosciences (Regier et al. 2025).This study used reach-scale modeled estimates of hyporheic aerobic respiration made by the River Corridor Model (Fang et al. 2020) and watershed characteristics across the Willamette and Yakima River basins to explore potential allometric scaling (i.e., power-law relationships between size and function) of cumulative hyporheic respiration across catchment-to-basin scales. Scaling was explored quantitatively via the R2, slope, and y-intercept of relationships between cumulative hyporheic respiration and watershed area, divided into hyporheic exchange flux (HEF) quantiles. We also explored relationships between allometric scaling and other watershed characteristics through linear regression, spatial patterns, and mutual information analyses. Our results also suggest variability of hyporheic respiration allometry for middle exchange flux quantiles, and in relation to land-cover. Our findings provide initial evidence that allometric scaling may be useful for predicting hyporheic biogeochemical dynamics across watersheds from reach to basin scales. This data package is associated with the GitHub repository found at https://github.com/peterregier/rc_wrb_yrb_scaling. The data package is organized into several key directories. The “data” folder contains multiple CSV files, including landscape heterogeneity, scaling analysis, and watershed boundary data. The “figures” folder has all figure files in both PDF and PNG formats. Core analysis scripts and figure generation scripts are in the “scripts” directory, systematically numbered for sequential execution. The root directory includes essential project files; please see the file ending in “flmd.csv” for a list and description of all files contained in this data package and the file ending in “dd.csv” for data dictionaries used to describe tabular column headers.

54 ENVIRONMENTAL SCIENCES↗

DOE OSTI2023

Bias Correction in Urban Building Energy Modeling for Chicago Using Machine Learning

Urban-scale building energy modeling (UBEM) holds promise for optimizing energy usage across extensive geographic regions. However, there is a recognized bias between simulated energy consumption and actual measured data. This study, based on building data from Chicago, delved into bias correction techniques for enhancing the accuracy of UBEM energy consumption estimates. Initially, the AutoBEM simulation yielded a normalized mean bias error (NMBE) of 1.1% and 51% of Coefficient of the Variation of the Root Mean Square Error (CVRMSE) after outlier exclusion. To address this, three bias correction methods were deployed: Average Mean Bias Error based bias correction, Quantile mapping bias correction, and Machine learning-based bias correction using Linear Regression and Random Forest models. Post-correction results exhibited marked improvement. The NMBE values were diminished to 0 for Average MBE-based, 0.36 for Quantile Mapping, and 0 for Machine Learning-based corrections. Concurrently, the CVRMSE values registered reductions from an original 51 to 50.8 for Quantile Mapping, and 38.56 for Machine Learning-based corrections, pointing towards the effectiveness of specific bias correction methods in refining the precision of UBEM energy predictions. Such accurate estimations are paramount for informed energy planning and urban policy-making.

Chowdhury, Shovan↗

DOE OSTI2024

Interpretable Net Load Forecasting Using Smooth Multiperiodic Features

We consider the problem of forecasting net load over a horizon such as one day, using a trailing window of past net load values as well as date and time. We focus on three variations on this problem: point forecasts, marginal quantile forecasts, and generating conditional samples of the future value. We propose a method that relies on linear regression using some custom engineered time-based features to capture multiple periodicities, such as daily, weekly, and seasonal, and their interactions. Our proposed models are readily interpretable, and rely on efficient and reliable convex optimization [1] to fit. We illustrate our method on four years worth of hourly net load data, comparing predictions made with various subsets of the features.

Ogut, Mehmet G↗

DOE OSTI2011

The effects of digital elevation model resolution on the calculation and predictions of topographic wetness indices.

One of the largest exports in the Southeast U.S. is forest products. Interest in biofuels using forest biomass has increased recently, leading to more research into better forest management BMPs. The USDA Forest Service, along with the Oak Ridge National Laboratory, University of Georgia and Oregon State University are researching the impacts of intensive forest management for biofuels on water quality and quantity at the Savannah River Site in South Carolina. Surface runoff of saturated areas, transporting excess nutrients and contaminants, is a potential water quality issue under investigation. Detailed maps of variable source areas and soil characteristics would therefore be helpful prior to treatment. The availability of remotely sensed and computed digital elevation models (DEMs) and spatial analysis tools make it easy to calculate terrain attributes. These terrain attributes can be used in models to predict saturated areas or other attributes in the landscape. With laser altimetry, an area can be flown to produce very high resolution data, and the resulting data can be resampled into any resolution of DEM desired. Additionally, there exist many maps that are in various resolutions of DEM, such as those acquired from the U.S. Geological Survey. Problems arise when using maps derived from different resolution DEMs. For example, saturated areas can be under or overestimated depending on the resolution used. The purpose of this study was to examine the effects of DEM resolution on the calculation of topographic wetness indices used to predict variable source areas of saturation, and to find the best resolutions to produce prediction maps of soil attributes like nitrogen, carbon, bulk density and soil texture for low-relief, humid-temperate forested hillslopes. Topographic wetness indices were calculated based on the derived terrain attributes, slope and specific catchment area, from five different DEM resolutions. The DEMs were resampled from LiDAR, which is a laser altimetry remote sensing method, obtained from the USDA Forest Service at Savannah River Site. The specific DEM resolutions were chosen because they are common grid cell sizes (10m, 30m, and 50m) used in mapping for management applications and in research. The finer resolutions (2m and 5m) were chosen for the purpose of determining how finer resolutions performed compared with coarser resolutions at predicting wetness and related soil attributes. The wetness indices were compared across DEMs and with each other in terms of quantile and distribution differences, then in terms of how well they each correlated with measured soil attributes. Spatial and non-spatial analyses were performed, and predictions using regression and geostatistics were examined for efficacy relative to each DEM resolution. Trends in the raw data and analysis results were also revealed.

09 BIOMASS FUELS↗

DOE OSTI2021

Predicting Intensive Care Unit Length of Stay and Mortality Using Patient Vital Signs: Machine Learning Model Development and Validation

Background: Patient monitoring is vital in all stages of care. In particular, intensive care unit (ICU) patient monitoring has the potential to reduce complications and morbidity, and to increase the quality of care by enabling hospitals to deliver higher-quality, cost-effective patient care, and improve the quality of medical services in the ICU. Objective: We here report the development and validation of ICU length of stay and mortality prediction models. The models will be used in an intelligent ICU patient monitoring module of an Intelligent Remote Patient Monitoring (IRPM) framework that monitors the health status of patients, and generates timely alerts, maneuver guidance, or reports when adverse medical conditions are predicted. Methods: We utilized the publicly available Medical Information Mart for Intensive Care (MIMIC) database to extract ICU stay data for adult patients to build two prediction models: one for mortality prediction and another for ICU length of stay. For the mortality model, we applied six commonly used machine learning (ML) binary classification algorithms for predicting the discharge status (survived or not). For the length of stay model, we applied the same six ML algorithms for binary classification using the median patient population ICU stay of 2.64 days. For the regression-based classification, we used two ML algorithms for predicting the number of days. We built two variations of each prediction model: one using 12 baseline demographic and vital sign features, and the other based on our proposed quantiles approach, in which we use 21 extra features engineered from the baseline vital sign features, including their modified means, standard deviations, and quantile percentages. Results: We could perform predictive modeling with minimal features while maintaining reasonable performance using the quantiles approach. The best accuracy achieved in the mortality model was approximately 89% using the random forest algorithm. The highest accuracy achieved in the length of stay model, based on the population median ICU stay (2.64 days), was approximately 65% using the random forest algorithm. Conclusions: The novelty in our approach is that we built models to predict ICU length of stay and mortality with reasonable accuracy based on a combination of ML and the quantiles approach that utilizes only vital signs available from the patient’s profile without the need to use any external features. This approach is based on feature engineering of the vital signs by including their modified means, standard deviations, and quantile percentages of the original features, which provided a richer dataset to achieve better predictive power in our models.

59 BASIC BIOLOGICAL SCIENCES↗

DOE OSTI2020

Quantile-dependent expressivity of postprandial lipemia

“Quantile-dependent expressivity” describes an effect of the genotype that depends upon the level of the phenotype (e.g., whether a subject’s triglycerides are high or low relative to its population distribution). Prior analyses suggest that the effect of a genetic risk score (GRS) on fasting plasma triglyceride levels increases with the percentile of the triglyceride distribution. Postprandial lipemia is well suited for testing quantile-dependent expressivity because it exposes each individual’s genotype to substantial increases in their plasma triglyceride concentrations. Ninety-seven published papers were identified that plotted mean triglyceride response vs. time and genotype, which were converted into quantitative data. Separately, for each published graph, standard least-squares regression analysis was used to compare the genotype differences at time t (dependent variable) to average triglyceride concentrations at time t (independent variable) to assess whether the genetic effect size increased in association with higher triglyceride concentrations and whether the phenomenon could explain purported genetic interactions with sex, diet, disease, BMI, and drugs.

59 BASIC BIOLOGICAL SCIENCES↗

NASA NTRS1982

Quantiles, parametric-select density estimation, and bi-information parameter estimators

A quantile-based approach to statistical analysis and probability modeling of data is presented which formulates statistical inference problems as functional inference problems in which the parameters to be estimated are density functions. Density estimators can be non-parametric (computed independently of model identified) or parametric-select (approximated by finite parametric models that can provide standard models whose fit can be tested). Exponential models and autoregressive models are approximating densities which can be justified as maximum entropy for respectively the entropy of a probability density and the entropy of a quantile density. Applications of these ideas are outlined to the problems of modeling: (1) univariate data; (2) bivariate data and tests for independence; and (3) two samples and likelihood ratios. It is proposed that bi-information estimation of a density function can be developed by analogy to the problem of identification of regression models.

Parzen, E.↗

DOE OSTI2021

Risk-Adaptive Experimental Design for High-Consequence Systems: LDRD Final Report

Constructing accurate statistical models of critical system responses typically requires an enormous amount of data from physical experiments or numerical simulations. Unfortunately, data generation is often expensive and time consuming. To streamline the data generation process, optimal experimental design determines the 'best' allocation of experiments with respect to a criterion that measures the ability to estimate some important aspect of an assumed statistical model. While optimal design has a vast literature, few researchers have developed design paradigms targeting tail statistics, such as quantiles. In this project, we tailored and extended traditional design paradigms to target distribution tails. Our approach included (i) the development of new optimality criteria to shape the distribution of prediction variances, (ii) the development of novel risk-adapted surrogate models that provably overestimate certain statistics including the probability of exceeding a threshold, and (iii) the asymptotic analysis of regression approaches that target tail statistics such as superquantile regression. To accompany our theoretical contributions, we released implementations of our methods for surrogate modeling and design of experiments in two complementary open source software packages, the ROL/OED Toolkit and PyApprox.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2025

wa-hls4ml and lui-gnn: A benchmark and GNN-based surrogate model for hls4ml resource and latency estimation

As machine learning (ML) increasingly serves as a tool for addressing real-time challenges in scientific applications, the development of advanced tooling has significantly reduced the time required to iterate on various designs. These advancements have solved major obstacles, but also exposed new challenges. For example, processes that were not previously considered bottlenecks, such as model synthesis, are now becoming limiting factors in the rapid iteration of designs. To reduce these emerging constraints, multiple efforts are being launched toward designing an ML-based surrogate model that estimates resource usage of synthesized accelerator architectures. This model would reduce the design iteration time, especially when designing within a set of given hardware constraints. This approach shows considerable potential, but as it stands, the effort is early and would benefit from coordination and standardization to assist future work as it emerges. We introduce wa-hls4ml, a benchmark for ML accelerator resource and latency estimation, and its corresponding initial dataset of more than 100,000 fully connected neural networks, all synthesized using hls4ml and targeting Xilinx FPGAs. In addition to the resource utilization and latency data provided, the dataset includes generated artifacts and log files for many of the synthesized neural networks, in order to support future research in ML-based code generation. The benchmark evaluates the performance of resource and latency predictors against several common ML model architectures, primarily originating from scientific domains, as exemplar models, as well as the average performance across a subset of the dataset. We measure the performance of a given predictor model through multiple metrics, including $R^2$ score and SMAPE on regression tasks, as well as inference time to further characterize the estimator under test. Additionally, we introduce the latency/utilization inference graph neural network (lui-gnn), a surrogate model that uses a graph neural network to represent input architectures in the form of a directed graph. This graph representation allows for a diverse set of model architectures to all be effectively handled by a surrogate model. We present the architecture and performance of the model, as evaluated by the new proposed benchmark, including SMAPE, $R^2$ score, and inference times, and find that lui-gnn generally predicts latency and utilization for the 75\% quantile within several percent of the synthesized resources on the synthetic test dataset, indicating that this approach of estimating resource and latency via a surrogate models has promise and warrants further research.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗