SEARCH · Engineering Papers

Results for “probability and statistical methods”

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 559 records · Page 31

All sources NASA NTRS DOE OSTI

NASA NTRS2003

Predicting protein functions from redundancies in large-scale protein interaction networks

Interpreting data from large-scale protein interaction experiments has been a challenging task because of the widespread presence of random false positives. Here, we present a network-based statistical algorithm that overcomes this difficulty and allows us to derive functions of unannotated proteins from large-scale interaction data. Our algorithm uses the insight that if two proteins share significantly larger number of common interaction partners than random, they have close functional associations. Analysis of publicly available data from Saccharomyces cerevisiae reveals >2,800 reliable functional associations, 29% of which involve at least one unannotated protein. By further analyzing these associations, we derive tentative functions for 81 unannotated proteins with high certainty. Our method is not overly sensitive to the false positives present in the data. Even after adding 50% randomly generated interactions to the measured data set, we are able to recover almost all (approximately 89%) of the original associations.

Proteins/chemistry/metabolism↗

DOE OSTI2025

Rarefied xenon flow in orificed hollow cathodes

A parametric study is conducted to quantify the effect of the keeper electrode geometry on the xenon neutral flow quantities within orificed hollow cathodes, prior to ignition. The keeper impinges directly on the flow out of the cathode orifice and its geometry influences the product between the pressure in the orifice–keeper region and the cathode-to-keeper distance. A representative cathode is simulated using the Direct Simulation Monte Carlo method. The numerical model is first validated with computational results from the literature. A parametric study is then conducted. Parameters include the cathode pressure–diameter in the range of 1–5 Torr cm and the following geometric ratios (and ranges): cathode orifice-to-inner radii (0.1–0.7), keeper orifice-to-cathode orifice radii (1–5), and keeper distance-to-cathode-orifice diameter (0.5–10). It is found that, if both keeper and cathode have identical orifice radii, the flow remains subsonic in the orifice-to-keeper region. In most cases, however, the flow becomes underexpanded and supersonic, and the static pressure within the orifice-to-keeper region is, on average, 4% that of the upstream pressure value. The orifice–keeper region pressure increases with either a decrease in the keeper orifice diameter or an increase in the distance between cathode and keeper, in agreement with literature data. Both trends are explained through conservation laws. A statistical study of numerical results reveals that the ratio of ignition-to-nominal mass flow rates has a most probable value of 50, which suggests that heaterless cathode ignition at a minimum DC voltage may be achieved by increasing the input mass flow rate by a factor of 50.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

NASA NTRS2020

Synergistic Use of Hyperspectral UV-Visible OMI and Broadband Meteorological Imager MODIS Data for a Merged Aerosol Product

The retrieval of optimal aerosol datasets by the synergistic use of hyperspectral ultraviolet(UV)–visible and broadband meteorological imager (MI) techniques was investigated. The Aura Ozone Monitoring Instrument (OMI) Level 1B (L1B) was used as a proxy for hyperspectral UV–visible instrument data to which the Geostationary Environment Monitoring Spectrometer (GEMS) aerosol algorithm was applied. Moderate-Resolution Imaging Spectroradiometer (MODIS) L1B and dark target aerosol Level 2 (L2) data were used with a broadband MI to take advantage of the consistent time gap between the MODIS and the OMI. First, the use of cloud mask information from the MI infrared (IR) channel was tested for synergy. High-spatial-resolution and IR channels of the MI helped mask cirrus and sub-pixel cloud contamination of GEMS aerosol, as clearly seen in aerosol optical depth (AOD) validation with Aerosol Robotic Network (AERONET) data. Second, dust aerosols were distinguished in the GEMS aerosol-type classification algorithm by calculating the total dust confidence index (TDCI) from MODIS L1B IR channels. Statistical analysis indicates that the Probability of Correct Detection (POCD) between the forward and inversion aerosol dust models (DS) was increased from 72% to 94% by use of the TDCI for GEMS aerosol-type classification, and updated aerosol types were then applied to the GEMS algorithm. Use of the TDCI for DS type classification in the GEMS retrieval procedure gave improved single-scattering albedo (SSA) values for absorbing fine pollution particles (BC) and DS aerosols. Aerosol layer height (ALH) retrieved from GEMS was compared with Cloud-Aerosol Lidar with Orthogonal Polarization (CALIOP) data, which provides high-resolution vertical aerosol profile information. The CALIOP ALH was calculated from total attenuated backscatter data at 1064 nm, which is identical to the definition of GEMS ALH. Application of the TDCI value reduced the median bias of GEMS ALH data slightly. The GEMS ALH bias approximates zero, especially for GEMS AOD values of>~0.4 and GEMS SSA values of<~0.95.Finally, the AOD products from the GEMS algorithm and MI were used in aerosol merging with the maximum-likelihood estimation method, based on a weighting factor derived from the standard deviation of the original AOD products. With the advantage of the UV–visible channel in retrieving aerosol properties over bright surfaces, the combined AOD products demonstrated better spatial data availability than the original AOD products, with comparable accuracy. Furthermore, pixel-level error analysis of GEMS AOD data indicates improvement through MI synergy.

aerosol↗

NASA NTRSDate not supplied

More Data Needed for Failure Rate Estimation, Validation, and Uncertainty Reduction

Current Environmental Control and Life Support System (ECLSS) development and test activities are not generating data fast enough to provide statistically-supportable precise Orbital Replacement Unit (ORU) failure rate estimates for future missions. Accurate and precise failure rate estimates are critical for missions beyond Low Earth Orbit (LEO) because current risk mitigation approaches – namely regular resupply and rapid abort capabilities – will not be available. Safe operations will depend on mission planners’ ability to accurately forecast spares demand and efficiently provide the necessary resources. However, even after more than a decade of operations on board the International Space Station (ISS), a significant amount of uncertainty remains in failure rate estimates. Uncertain or inaccurate failure rates result in increased risk and spares mass for future missions. A Bayesian failure rate estimation approach, such as the one currently implemented by the ISS Program, can help reduce uncertainty by incorporating engineering judgement into failure rate estimates. However, experience on the ISS and with other complex systems shows that these prior failure rate estimates are often inaccurate. In addition, prior failure rate estimates are typically point values; some level of uncertainty must be added to convert these into probability distributions for Bayesian updating, and there are several potential methods for doing so. Due to the low rate of data collection, these subjective (and often inaccurate) prior estimates currently have a strong influence on the end result. This paper examines the challenges associated with failure rate estimation, validation, and uncertainty reduction in the context of ECLSS development for beyond-LEO missions. A variety of techniques for generating and updating Bayesian priors are discussed and evaluated using both real-world and simulated data. Potential solutions for improving failure rate estimation, including testing additional units, are analyzed and discussed, and a set of recommendations are provided for next-generation system development activities.

Supportability↗

DOE OSTI2025

Deriving the Landauer Principle From the Quantum Shannon Entropy

We derive an expression to determine the equilibrium probability distribution of a quantum state in contact with a noisy thermal environment that formally separates contributions from quantum and classical forms of probabilistic uncertainty. A statistical mechanical interpretation of this probability distribution enables us to derive an expression for the minimum free energy costs for arbitrary (reversible or irreversible) quantum state changes. In conclusion, based on this derivation, we demonstrate that–in contrast to classical systems–the free energy required to erase or reset a qubit depends sensitively on both the fidelity of the target state and on the physical properties of the environment, such as the number of quantum bath states, due primarily to the entropic effects of system-bath entanglement.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

NASA NTRS2008

Potential Use of a Bayesian Network for Discriminating Flash Type from Future GOES-R Geostationary Lightning Mapper (GLM) data

Continuous monitoring of the ratio of cloud flashes to ground flashes may provide a better understanding of thunderstorm dynamics, intensification, and evolution, and it may be useful in severe weather warning. The National Lighting Detection Network TM (NLDN) senses ground flashes with exceptional detection efficiency and accuracy over most of the continental United States. A proposed Geostationary Lightning Mapper (GLM) aboard the Geostationary Operational Environmental Satellite (GOES-R) will look at the western hemisphere, and among the lightning data products to be made available will be the fundamental optical flash parameters for both cloud and ground flashes: radiance, area, duration, number of optical groups, and number of optical events. Previous studies have demonstrated that the optical flash parameter statistics of ground and cloud lightning, which are observable from space, are significantly different. This study investigates a Bayesian network methodology for discriminating lightning flash type (ground or cloud) using the lightning optical data and ancillary GOES-R data. A Directed Acyclic Graph (DAG) is set up with lightning as a "root" and data observed by GLM as the "leaves." This allows for a direct calculation of the joint probability distribution function for the lighting type and radiance, area, etc. Initially, the conditional probabilities that will be required can be estimated from the Lightning Imaging Sensor (LIS) and the Optical Transient Detector (OTD) together with NLDN data. Directly manipulating the joint distribution will yield the conditional probability that a lightning flash is a ground flash given the evidence, which consists of the observed lightning optical data [and possibly cloud data retrieved from the GOES-R Advanced Baseline Imager (ABI) in a more mature Bayesian network configuration]. Later, actual GLM and NLDN data can be used to refine the estimates of the conditional probabilities used in the model; i.e., the Bayesian network is a learning network. Methods for efficient calculation of the conditional probabilities (e.g., an algorithm using junction trees), finding data conflicts, goodness of fit, and dealing with missing data will also be addressed.

Solakiewiz, Richard↗

DOE OSTI2020

Quantified limits of the nuclear landscape

The chart of the nuclides is limited by particle drip lines beyond which nuclear stability to proton or neutron emission is lost. Predicting the range of particle-bound isotopes poses an appreciable challenge for nuclear theory as it involves extreme extrapolations of nuclear masses well beyond the regions where experimental information is available. Still, quantified extrapolations are crucial for a wide variety of applications, including the modeling of stellar nucleosynthesis. We use microscopic nuclear global mass models, current mass data, and Bayesian methodology to provide quantified predictions of proton and neutron separation energies as well as Bayesian probabilities of existence throughout the nuclear landscape all the way to the particle drip lines. Here, we apply nuclear density functional theory with several energy density functionals. We also consider two global mass models often used in astrophysical nucleosynthesis simulations. To account for uncertainties, Bayesian Gaussian processes are trained on the separation-energy residuals for each individual model, and the resulting predictions are combined via Bayesian model averaging. This framework allows to account for systematic and statistical uncertainties and propagate them to extrapolative predictions. We establish and characterize the drip-line regions where the probability that the nucleus is particle- bound decreases from 1 to 0. In these regions, we provide quantified predictions for one- and two-nucleon separation energies. According to our Bayesian model averaging analysis, 7759 nuclei with Z ≤ 119 have a probability of existence ≥ 0.5. The extrapolation results obtained in this study will be put through stringent tests when new experimental information on existence and masses of exotic nuclei becomes available. In this respect, the quantified landscape of nuclear existence obtained in this study should be viewed as a dynamical prediction that will be fine-tuned when new experimental information and improved global mass models become available.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

DOE OSTI2020

Parametrizing the detector response with neural networks

In high energy physics, characterizing the response of a detector to radiation is one of the most important and fundamental experimental tasks. In many cases, this task is accomplished by parameterizing summary statistics of the full detector response probability density. The parameterized detector response can then be used for calibration as well as for directly improving physics analysis sensitivity. This paper reports on how to parameterize summary statistics of the detector response using neural networks. Specifically, neural networks are powerful tools for incorporating multidimensional data and the loss function used during training determines which summary statistic is learned. One common summary statistic that has not been combined with deep learning (as far as the authors are aware) is the mode. A neural network-based approach to mode learning is proposed and empirically demonstrated in the context of high energy jet calibrations. Altogether, the neural network-based toolkit for detector response parameterization can enhance the utility of data collected at high energy physics experiments and beyond.

47 OTHER INSTRUMENTATION↗

DOE OSTI2023

The entropy of galaxy spectra: how much information is encoded?

Abstract The inverse problem of extracting the stellar population content of galaxy spectra is analysed here from a basic standpoint based on information theory. By interpreting spectra as probability distribution functions, we find that galaxy spectra have high entropy, thus leading to a rather low effective information content. The highest variation in entropy is unsurprisingly found in regions that have been well studied for decades with the conventional approach. We target a set of six spectral regions that show the highest variation in entropy – the 4000 Å break being the most informative one. As a test case with real data, we measure the entropy of a set of high-quality spectra from the Sloan Digital Sky Survey, and contrast entropy-based results with the traditional method based on line strengths. The data are classified into star-forming (SF), quiescent (Q), and active galactic nucleus (AGN) galaxies, and show – independently of any physical model – that AGN spectra can be interpreted as a transition between SF and Q galaxies, with SF galaxies featuring a more diverse variation in entropy. The high level of entanglement complicates the determination of population parameters in a robust, unbiased way, and affects traditional methods that compare models with observations, as well as machine learning (especially deep learning) algorithms that rely on the statistical properties of the data to assess the variations among spectra. Entropy provides a new avenue to improve population synthesis models so that they give a more faithful representation of real galaxy spectra.

Ferreras, Ignacio (ORCID:0000000345843127)↗

NASA NTRSDate not supplied

Artificial Intelligence (AI) Methods for Automating the Impact Tool Evidence Library

INTRODUCTION: The development of the Evidence Library for use with the IMPACT (Informing Mission Planning via Analysis of Complex Tradespaces) probability risk assessment tool involved a multilayered, time intensive process of data collection and analysis by subject matter experts from the Exploration Medical Capability (ExMC) Element Clinical and Science Team to produce clinical findings forms (CliFFs) for 120 medical conditions. Artificial Intelligence Large Language Models (LLMs) can be leveraged to facilitate this process, thus reducing labor and time. TOPIC: CliFFs contain information about medical conditions as they pertain to spaceflight. This includes condition definitions, incidence data, crew task impairment estimates caused by conditions, treatment protocols and references to literature used for gathering condition evidence. Guided by the Evidence Library Methods document and the CliFF development instructions, a team has leveraged Microsoft Azure AI services and open-source documentation to construct an AI-assisted automated pipeline for CliFF development. This process is designed to search, retrieve, and evaluate the applicable data, and ultimately generate a completed CliFF. The LLM evaluates the relevance of each of the source materials to spaceflight, either as direct evidence or as an analog. The model extracts keywords and generates brief summaries to enhance search and retrieval in later stages of CliFF development. For instance, it can calculate epidemiological statistical data, such as incidence rates and the likelihood of best or worst-case scenarios. APPLICATION: Large Language Models (LLMs) can efficiently summarize large amounts of text. Leveraging this technology will automate data retrieval and evidence gathering for medical databases, like the IMPACT tool, by aiding in the labor-intensive process of analyzing large bodies of literature and organizing it into a formatted document like a CliFF. This added efficiency will enable expeditious expansion of the Evidence Library with additional medical conditions and update previous CLiFFs as new technology becomes available.

Ali Al↗

DOE OSTI2020

Convergence acceleration of Monte Carlo many-body perturbation methods by direct sampling

In the Monte Carlo many-body perturbation (MC-MP) method, the conventional correlation-correction formula, which is a long sum of products of low-dimensional integrals, is first recast into a short sum of high-dimensional integrals over electron-pair and imaginary-time coordinates. These high-dimensional integrals are then evaluated by the Monte Carlo method with random coordinates generated by the Metropolis–Hasting algorithm according to a suitable distribution. The latter algorithm, while advantageous in its ability to sample nearly any distribution, introduces autocorrelation in sampled coordinates, which in turn increases the statistical uncertainty of the integrals and thus the computational cost. It also involves wasteful rejected moves and an initial “burn-in” step as well as displays hysteresis. Here, an algorithm is proposed that directly produces a random sequence of electron-pair coordinates for the same distribution used in the MC-MP method, which is free from autocorrelation, rejected moves, a burn-in step, or hysteresis. Furthermore, this direct-sampling algorithm is shown to accelerate second- (MC-MP2) and third-order Monte Carlo many-body perturbation (MC-MP3) calculations by up to 222% and 38%, respectively.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

NASA NTRS2004

Bounding the risk of crew loss following orbital debris penetration of the International Space Station at assembly stages 1J and 1E

Orbital debris impacts on the International Space Station occur frequently. To date, none of the impacting particles has been large enough to penetrate manned pressurized volumes. We used the Manned Spacecraft Crew Survivability code to evaluate the risk to crew of penetrations of pressurized modules at two assembly stages: after Flight 1J, when the pressurized elements of Kibo, the Japanese Experiment Module, are present, and after Flight 1E, when the European Columbus Module is present. Our code is a Monte-Carlo simulation of impacts on the Station that considers several potential event types that could lead to crew loss. Among the statistics tabulated by the program is the probability of death of one or more crew members in the event of a penetration, expressed as the risk factor, R. This risk factor is dependent on details of crew operations during both ordinary circumstances and decompression emergencies, as well as on details of internal module configurations. We conducted trade studies considering these procedure and configuration details to determine the bounds on R at the 1J and 1E stages in the assembly sequence. Here we compare the R-factor bounds, and procedures could that reduce R at these stages. Published by Elsevier Ltd on behalf of COSPAR.

Models, Theoretical↗

DOE OSTI2021

Bayesian model averaging for analysis of lattice field theory results

Statistical modeling is a key component in the extraction of physical results from lattice field theory calculations. Although the general models used are often strongly motivated by physics, many model variations can frequently be considered for the same lattice data. Model averaging, which amounts to a probability-weighted average over all model variations, can incorporate systematic errors associated with model choice without being overly conservative. We discuss the framework of model averaging from the perspective of Bayesian statistics, and give useful formulae and approximations for the particular case of least-squares fitting, commonly used in modeling lattice results. In addition, we frame the common problem of data subset selection (e.g. choice of minimum and maximum time separation for fitting a two-point correlation function) as a model selection problem and study model averaging as a straightforward alternative to manual selection of fit ranges. Numerical examples involving both mock and real lattice data are given.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

NASA NTRS2008

Use of High-Resolution WRF Simulations to Forecast Lightning Threat

Recent observational studies have confirmed the existence of a robust statistical relationship between lightning flash rates and the amount of large precipitating ice hydrometeors aloft in storms. This relationship is exploited, in conjunction with the capabilities of cloud-resolving forecast models such as WRF, to forecast explicitly the threat of lightning from convective storms using selected output fields from the model forecasts. The simulated vertical flux of graupel at -15C and the shape of the simulated reflectivity profile are tested in this study as proxies for charge separation processes and their associated lightning risk. Our lightning forecast method differs from others in that it is entirely based on high-resolution simulation output, without reliance on any climatological data. short [6-8 h) simulations are conducted for a number of case studies for which three-dmmensional lightning validation data from the North Alabama Lightning Mapping Array are available. Experiments indicate that initialization of the WRF model on a 2 km grid using Eta boundary conditions, Doppler radar radial velocity fields, and METAR and ACARS data y&eld satisfactory simulations. __nalyses of the lightning threat fields suggests that both the graupel flux and reflectivity profile approaches, when properly calibrated, can yield reasonable lightning threat forecasts, although an ensemble approach is probably desirable in order to reduce the tendency for misplacement of modeled storms to hurt the accuracy of the forecasts. Our lightning threat forecasts are also compared to other more traditional means of forecasting thunderstorms, such as those based on inspection of the convective available potential energy field.

McCaul, E. W., Jr.↗

DOE OSTI2024

Assessing correlated truncation errors in modern nucleon-nucleon potentials

We test the BUQEYE model of correlated effective field theory (EFT) truncation errors on Reinert, Krebs, and Epelbaum's semilocal momentum-space implementation of the chiral EFT (𝜒⁢EFT ) expansion of the nucleon-nucleon (NN) potential. This Bayesian model hypothesizes that dimensionless coefficient functions extracted from the order-by-order corrections to NN observables can be treated as draws from a Gaussian process (GP). We combine a variety of graphical and statistical diagnostics to assess when predicted observables have a 𝜒⁢EFT convergence pattern consistent with the hypothesized GP statistical model. Our conclusions are that, first, the BUQEYE model is generally applicable to the potential investigated here, which enables statistically principled estimates of the impact of higher EFT orders on observables. Second, parameters defining the extracted coefficients such as the expansion parameter 𝑄 must be well chosen for the coefficients to exhibit a regular convergence pattern—a property we exploit to obtain posterior distributions for such quantities. Third, the assumption of GP stationarity across lab energy and scattering angle is not generally met; this necessitates adjustments in future work. We provide a workflow and interpretive guide for our analysis framework, and show what can be inferred about probability distributions for 𝑄, the EFT breakdown scale Λ 𝑏 , the scale associated with soft physics in the 𝜒⁢EFT potential 𝑚 eff , and the GP hyperparameters. All our results can be reproduced using a publicly available Jupyter notebook, which can be straightforwardly modified to analyze other 𝜒⁢EFT NN potentials.

Bayesian methods↗

NASA NTRS2020

Risk and Performance Assessment of Generic Mission Architectures: Showcasing the Artemis Mission

A has initiated a strong push to return face. In this work, we astronaut assess performance and risk for proposed mission architectures using a new Mission Architecture Risk Assessment (MARA) tool. The MARA tool can produce statistics about the availability of components and overall performance of the mission considering potential failures of any of its components. In a Monte Carlo approach, the tool repeats the mission simulation multiple times while a random generator lets modules fail according to their failure rates. The results provide statistically meaningful insights into the overall performance of the chosen architecture. A given mission architecture can be freely replicated in the tool, with the mission timeline and basic characteristics of employed mission modules (habitats, rovers, power generation units, etc.) specified in a configuration file. Crucially, failure rates for each module need to be known or estimated. The tool performs an event-driven simulation of the mission and accounts for random failure events. Failed modules can be repaired, which takes crew time but restores operations. In addition to tracking individual modules, MARA can assess the availability of predefined functions throughout the mission. For instance, the function of resource collection would require a rover to collect the resources, a power generation unit to charge the rover, and a resource processing module. Together, the modules that are required for a given function are called a functional group. Similarly, we can assess how much crew time is available to achieve a mission benefit (e.g. research, building a base, etc) as opposed to spending crew time on repairs. Here we employ the method on the proposed NASA Artemis mission. Artemis aims to return United States astronauts to the lunar surface by 2024. Results provide insights into mission failure probabilities, up- and downtime for individual modules and crew-time resources spent on the repair of failed modules. The tool also allows us to tweak the mission architecture in order to find setups that produce more favorable mission performance. As such, the tool can be an aid in improving the mission architect abling cost-benefit analysis for mission improvement.

Rumpf, Clemens M.↗

DOE OSTI2022

Kinetic theory of particle-in-cell simulation plasma and the ensemble averaging technique

Abstract We derive the kinetic theory of fluctuations in physically and numerically stable particle-in-cell (PIC) simulations of electrostatic plasmas. The starting point is the single-time correlation at the start of the simulation between the statistical fluctuations of the weighted densities of macroparticle centers in the plasma particle phase-space. The fluctuations are associated with different initial conditions, typically due to the random initial conditions (in velocity space) of the macroparticles/simulation plasma, assigned according to their initial distribution of probability. The single-time correlations at all time steps and in each spatial grid cell are then determined from the Laplace–Fourier transforms of the discretized Klimontovich-like equation for the macroparticles and Maxwell’s equations for the fields, as computed by modern PIC codes. We recover the expressions for the electrostatic field and the plasma particle density fluctuation autocorrelation spectra as well as the kinetic equations describing the average evolution of PIC-simulated plasma particles, first derived by Langdon (1970b Proc. 4th Conf. Numerical Simulation of Plasmas ) using a test macroparticle approach perturbing a discretized Vlasovian plasma and then averaging the obtained physical quantity over the initial macroparticle velocity distribution. We generalize and extend these results to the modern algorithms in PIC codes using arbitrary macroparticle weights. Analytical estimates of statistical fluctuation amplitudes are derived as a function of the plasma simulation parameters, using the central limit theorem in the limit of a large number of macroparticles per cell. The theory is then used to analyze the ensemble averaging technique of PIC simulations where statistical averages are performed over ensembles of PIC simulations, modeling the same plasma physics problem but using different statistical realizations of the initial distribution functions of the macroparticles. This method is illustrated by linear Landau damping uncovering (from noise, which is usually considered numerical) the physical fluctuations driven by a single small amplitude electrostatic wave perturbing a PIC simulation plasma in equilibrium.

fluctuations correlations↗

DOE OSTI2025

Combining Machine Learning and Comparative Effectiveness Methodology to Study Primary Care Pharmacotherapy Pathways for Veterans With Depression

Our objective is to demonstrate an innovative method combining machine learning with comparative effectiveness research techniques and to investigate a hitherto unstudied question about the effectiveness of common prescribing patterns. For Operation Enduring Freedom/Operation Iraqi Freedom veterans with major depressive disorder, we generate pharmacotherapy pathways (of antidepressants) using process mining and machine learning. We select the medication episodes that were started at subtherapeutic doses by the first assigned primary care physician and observe the paths that those medication episodes follow. Using 2-stage least squares, we test the effectiveness of starting at a low dose and staying low for longer versus ramping up fast while balancing observable and unobservable characteristics of patients and providers through instrumental variables. We leverage predetermined provider practice patterns as instruments. We collected outpatient pharmacy data for selective serotonin reuptake inhibitors and selective norepinephrine reuptake inhibitors, patient and provider characteristics (as control variables), and the instruments for our cohort. All data were extracted for the period between 2006 and 2020. There is a statistically significant positive effect (0.68, 95% CI 0.11–1.25) of “ramping up fast” on engagement in care. When we examine the effect of “ramping up slow”, we see an insignificant negative impact on engagement in care (−0.82, 95% CI −1.89 to 0.25). As expected, the probability of drop-out also seems to have a negative effect on engagement in care (−0.39, 95% CI −0.94 to 0.17). We further validate these results by testing with medication possession ratios calculated periodically as an alternative engagement in care metric. Our findings contradict the “Start low, go slow” adage, indicating that ramping up the dose of an antidepressant faster has a significantly positive effect on engagement in care for our population.

60 APPLIED LIFE SCIENCES↗