Search NASA⌕ Search

SEARCH · Search NASA

Results for “statistical comparison”

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 1,063 records · Page 59

A study of environmental characterization of conventional and advanced aluminum alloys for selection and design. Phase 2: The breaking load test method

A technique is demonstrated for accelerated stress corrosion testing of high strength aluminum alloys. The method offers better precision and shorter exposure times than traditional pass fail procedures. The approach uses data from tension tests performed on replicate groups of smooth specimens after various lengths of exposure to static stress. The breaking strength measures degradation in the test specimen load carrying ability due to the environmental attack. Analysis of breaking load data by extreme value statistics enables the calculation of survival probabilities and a statistically defined threshold stress applicable to the specific test conditions. A fracture mechanics model is given which quantifies depth of attack in the stress corroded specimen by an effective flaw size calculated from the breaking stress and the material strength and fracture toughness properties. Comparisons are made with experimental results from three tempers of 7075 alloy plate tested by the breaking load method and by traditional tests of statistically loaded smooth tension bars and conventional precracked specimens.

Sprowls, D. O.↗

Comparison of threshold selection methods for microarray gene co-expression matrices

Background: Network and clustering analyses of microarray co-expression correlation data often require application of a threshold to discard small correlations, thus reducing computational demands and decreasing the number of uninformative correlations. This study investigated threshold selection in the context of combinatorial network analysis of transcriptome data. Findings: Six conceptually diverse methods - based on number of maximal cliques, correlation of control spots with expressed genes, top 1% of correlations, spectral graph clustering, Bonferroni correction of p-values, and statistical power - were used to estimate a correlation threshold for three time-series microarray datasets. The validity of thresholds was tested by comparison to thresholds derived from Gene Ontology information. Stability and reliability of the best methods were evaluated with block bootstrapping. Two threshold methods, number of maximal cliques and spectral graph, used information in the correlation matrix structure and performed well in terms of stability. Comparison to Gene Ontology found thresholds from number of maximal cliques extracted from a co-expression matrix were the most biologically valid. Approaches to improve both methods were suggested. Conclusion: Threshold selection approaches based on network structure of gene relationships gave thresholds with greater relevance to curated biological relationships than approaches based on statistical pair-wise relationships.

59 BASIC BIOLOGICAL SCIENCES↗

Studies regarding the quality of numerical weather forecasts of the WRF model integrated at high-resolutions for the Romanian territory

Our paper aims to evaluate the quality of high-resolution weather forecasts from the Weather Research and Forecasting (WRF) numerical weather prediction model. The lateral and boundary conditions were obtained from the numerical output of the Consortium for Small-scale Modeling (COSMO) model at 7 km horizontal resolution. Furthermore, the WRF model was run for January and July 2013 at two horizontal resolutions (3 and 1 km). The numerical forecasts of the WRF model were evaluated using different statistical scores for 2 m temperature and 10 m wind speed. Our results showed a tendency of the WRF model to overestimate the values of the analyzed parameters in comparison to observations.

54 ENVIRONMENTAL SCIENCES↗

Fluid and gyrokinetic turbulence in open field-line, helical plasmas

Two-fluid Braginskii codes have simulated open-field line turbulence for over a decade, and only recently, it has become possible to study these systems with continuum gyrokinetic codes. This work presents a first-of-its-kind comparison between fluid and (long-wavelength) gyrokinetic models in open field-lines, using the GDB and Gkeyll codes to simulate interchange turbulence in the Helimak device at the University of Texas [T. N. Bernard et al., Phys. Plasmas 26, 042301 (2019)]. Partial agreement is attained in a number of diagnostic channels when the GDB sources and sheath boundary conditions (BCs) are selected carefully, especially the heat-flux BCs that can drastically alter the temperature. The radial profile of the fluctuation levels is qualitatively similar and quantitatively comparable on the low-field side, although statistics such as moments of the probability density function and the high-frequency spectrum show greater differences. This comparison indicates areas for future improvement in both simulations, such as sheath BCs, and improvements in GDB like particle conservation and spatially varying thermal conductivity, in order to achieve better fluid-gyrokinetic agreement and increase fidelity when simulating experiments.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Statistical Calibration and Validation of a Homogeneous Ventilated Wall-Interference Correction Method for the National Transonic Facility

Wind tunnel experiments will continue to be a primary source of validation data for many types of mathematical and computational models in the aerospace industry. The increased emphasis on accuracy of data acquired from these facilities requires understanding of the uncertainty of not only the measurement data but also any correction applied to the data. One of the largest and most critical corrections made to these data is due to wall interference. In an effort to understand the accuracy and suitability of these corrections, a statistical validation process for wall interference correction methods has been developed. This process is based on the use of independent cases which, after correction, are expected to produce the same result. Comparison of these independent cases with respect to the uncertainty in the correction process establishes a domain of applicability based on the capability of the method to provide reasonable corrections with respect to customer accuracy requirements. The statistical validation method was applied to the version of the Transonic Wall Interference Correction System (TWICS) recently implemented in the National Transonic Facility at NASA Langley Research Center. The TWICS code generates corrections for solid and slotted wall interference in the model pitch plane based on boundary pressure measurements. Before validation could be performed on this method, it was necessary to calibrate the ventilated wall boundary condition parameters. Discrimination comparisons are used to determine the most representative of three linear boundary condition models which have historically been used to represent longitudinally slotted test section walls. Of the three linear boundary condition models implemented for ventilated walls, the general slotted wall model was the most representative of the data. The TWICS code using the calibrated general slotted wall model was found to be valid to within the process uncertainty for test section Mach numbers less than or equal to 0.60. The scatter among the mean corrected results of the bodies of revolution validation cases was within one count of drag on a typical transport aircraft configuration for Mach numbers at or below 0.80 and two counts of drag for Mach numbers at or below 0.90.

Walker, Eric Lee↗

Statistical Calibration and Validation of a Homogeneous Ventilated Wall-Interference Correction Method for the National Transonic Facility

Wind tunnel experiments will continue to be a primary source of validation data for many types of mathematical and computational models in the aerospace industry. The increased emphasis on accuracy of data acquired from these facilities requires understanding of the uncertainty of not only the measurement data but also any correction applied to the data. One of the largest and most critical corrections made to these data is due to wall interference. In an effort to understand the accuracy and suitability of these corrections, a statistical validation process for wall interference correction methods has been developed. This process is based on the use of independent cases which, after correction, are expected to produce the same result. Comparison of these independent cases with respect to the uncertainty in the correction process establishes a domain of applicability based on the capability of the method to provide reasonable corrections with respect to customer accuracy requirements. The statistical validation method was applied to the version of the Transonic Wall Interference Correction System (TWICS) recently implemented in the National Transonic Facility at NASA Langley Research Center. The TWICS code generates corrections for solid and slotted wall interference in the model pitch plane based on boundary pressure measurements. Before validation could be performed on this method, it was necessary to calibrate the ventilated wall boundary condition parameters. Discrimination comparisons are used to determine the most representative of three linear boundary condition models which have historically been used to represent longitudinally slotted test section walls. Of the three linear boundary condition models implemented for ventilated walls, the general slotted wall model was the most representative of the data. The TWICS code using the calibrated general slotted wall model was found to be valid to within the process uncertainty for test section Mach numbers less than or equal to 0.60. The scatter among the mean corrected results of the bodies of revolution validation cases was within one count of drag on a typical transport aircraft configuration for Mach numbers at or below 0.80 and two counts of drag for Mach numbers at or below 0.90.

Walker, Eric L.↗

Cosmic rays near the heliospheric current sheet. II - An ensemble approach to comparing theory and observation

A quantitative comparison is carried out between theoretical predictions and observations of the intensity of galactic cosmic rays near the interplanetary current sheet. Model calculations are compared with a statistical analysis of observations of galactic cosmic rays at the earth and the simultaneous position of the current sheet. Since the observations were made over a period of several years, the inclination of the current sheet varied considerably, and comparison with any one model calculation referring to a given inclination would not be appropriate. An ensemble of different current sheet inclinations are used to generate expected values with the model in order to make the analysis of the computations approximate the method used to analyze the data. Agreement is found between theory and observation at energies of the order of a few GeV.

Jokipii, J. R.↗

Comparison of numerical weather prediction based deterministic and probabilistic wind resource assessment methods

Numerical weather prediction (NWP) models have been widely used for wind resource assessment. Model runs with higher spatial resolution are generally more accurate, yet extremely computational expensive. An alternative approach is to use data generated by a low resolution NWP model, in conjunction with statistical methods. In order to analyze the accuracy and computational efficiency of different types of NWP-based wind resource assessment methods, this paper performs a comparison of three deterministic and probabilistic NWP-based wind resource assessment methodologies: (i) a coarse resolution (0.5 degrees x 0.67 degrees) global reanalysis data set, the Modern-Era Retrospective Analysis for Research and Applications (MERRA); (ii) an analog ensemble methodology based on the MERRA, which provides both deterministic and probabilistic predictions; and (iii) a fine resolution (2-km) NWP data set, the Wind Integration National Dataset (WIND) Toolkit, based on the Weather Research and Forecasting model. Results show that: (i) as expected, the analog ensemble and WIND Toolkit perform significantly better than MERRA confirming their ability to downscale coarse estimates; (ii) the analog ensemble provides the best estimate of the multi-year wind distribution at seven of the nine sites, while the WIND Toolkit is the best at one site; (iii) the WIND Toolkit is more accurate in estimating the distribution of hourly wind speed differences, which characterizes the wind variability, at five of the available sites, with the analog ensemble being best at the remaining four locations; and (iv) the analog ensemble computational cost is negligible, whereas the WIND Toolkit requires large computational resources. Future efforts could focus on the combination of the analog ensemble with intermediate resolution (e.g., 10-15 km) NWP estimates, to considerably reduce the computational burden, while providing accurate deterministic estimates and reliable probabilistic assessments.

17 WIND ENERGY↗

Assessment of different coarse graining strategies to simulate polydisperse gas-solids flow

Continuum methods require the additional development of solids stress closures for polydisperse powders based on complex kinetic theories that are non-trivial to develop, code, and numerically converge for the wide range of fluidization regimes from very dilute to dense/frictional flow limit. On the other hand, it is straightforward to model the flow of polydisperse granular materials by treating particles as discrete rigid bodies that are tracked following simple physical laws of motion. The coarsening of these discrete methods by lumping several particles in a parcel alleviates the significant computational cost associated with these discrete methods while introducing some inaccuracies in the numerical results. In this research, we explore two different coarse graining methods that can be applied to polydisperse powders, namely the same statistic weight method (SSW) and the same size parcel method (SSP), and assess their accuracy by comparison with the finest simulation results obtained with a discrete element method (DEM). For Geldart group B powders fluidized at a relative low superficial velocity, the numerical results indicate that the SSW is more accurate than the SSP method. For type A powders fluidized at relatively high velocity, these two methods predict similar results. Interestingly, up to four times increase in the speed of simulation of the SSP method was obtained because the original polydisperse powder is scaled to a mono-disperse system in terms of particle-particle collision. Lastly, these results suggest that the SSP method is more favorable for the simulation of fluidized beds due to its accuracy and efficiency while the SSW method may be used for granular flow and dense fluidized bed systems where capturing the size segregation of particles due to collision is important.

42 ENGINEERING↗

Stage Presentation, Care Patterns, and Treatment Outcomes for Squamous Cell Carcinoma of the Penis

Purpose: Penile squamous cell carcinoma (SCC) is a rare entity, with few published series on outcomes. We evaluated the stage distributions and outcomes for surgery and radiation therapy in a U.S. population database. Methods and Materials: Subjects with SCC of the penis were identified using the National Cancer Institute Surveillance, Epidemiology and End Results (SEER) Program database between 1988 and 2006. Descriptive statistics were performed, and cause-specific survival (CSS) was estimated using Kaplan-Meier analysis. Comparisons of treatment modalities were analyzed using multivariate Cox regression. Subjects were staged using American Joint Committee on Cancer, sixth edition, criteria. Results: There were 2458 subjects identified. The median age was 66.8 years (range, 17-102 years). Grade 2 disease was present in 94.5% of cases. T1, T2, T3, T4, and Tx disease was present in 64.8%, 17.1%, 9.5%, 2.1%, and 6.5% of cases, respectively. N0, N1, N2, N3, and Nx disease was noted in 61.6%, 6.9%, 4.0%, 3.7%, and 23.8% of cases, respectively. M1 disease was noted in 2.5% of subjects. Individuals of white ethnicity accounted for 85.1% of cases. Lymphadenectomy was performed in 16.7% of cases. The CSS for all patients at 5 and 10 years was 80.8% and 78.6%. By multivariable analysis grades 2 and 3 disease, T3 stage, and positive lymph nodes were adverse prognostic factors for CSS. Conclusion: SCC of the penis often presents as early-stage T1, N0, M0, grade 1, or grade 2 disease. The majority of patients identified were treated with surgery, and only a small fraction of patients received radiation therapy alone or as adjuvant therapy.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

An evaluation of the one-dimensional turbulence model: Comparison with direct numerical simulations of CO/H2 jets with extinction and reignition

A variant of the one-dimensional turbulence (ODT) model formulated in an Eulerian reference frame is applied to a planar nonpremixed turbulent jet flame and results from the model prediction are compared with DNS data. The model employed herein solves the full set of conservation equations for mass, momentum, energy, and species on a one-dimensional domain corresponding to the transverse jet direction. The effects of turbulent mixing are modeled via a stochastic process, while the full range of diffusive-reactive length and time scales are resolved directly on the one-dimensional domain. A detailed chemical mechanism consisting of 11 species and 21 reactions and mixture-averaged transport is used in this study (consistent with DNS simulations). Comparisons between the model and DNS data in physical and state space are shown, including conditional statistics. Results indicate that the model accurately reproduces the DNS data set. Turbulence-chemistry interactions, including trends for extinction and reignition, are captured by the model. Differences observed between model prediction and data are the result of early excess extinction observed in the model. The reasons for the early extinction are discussed within the model context.

Energy & Fuels↗

Mixing and detonation structure in a rotating detonation engine with an axial air inlet

High-fidelity simulations of an experimental rotating detonation engine with an axial air inlet were conducted. The system operated with hydrogen as fuel at globally stoichiometric conditions. Instantaneous data showed that the detonation front is highly corrugated, and is considerably weaker than an ideal Chapman–Jouguet wave. Regions of deflagration are present ahead of the wave, caused by mixing with product gases from the previous cycle, as well as the injector recovery process. It is found that as the post-detonation high pressure flow expands, the injectors recover unsteadily, leading to a transient mixing process ahead of the next cycle. The resulting flow structure not only promotes mixing between product and reactant gases, but also increases likelihood of autoignition. These results show that the detonation process is very sensitive to injector design and the transient behavior during the detonation cycle. Phase-averaged statistics and conditionally averaged data are used to understand the overall reaction structure. As a result, comparisons with available experimental data on this configuration show remarkable good agreement of the predicted reacting flow structure.

42 ENGINEERING↗

Short-lead seasonal precipitation forecast in northeastern Brazil using an ensemble of artificial neural networks

This study assesses the deterministic and probabilistic forecasting skill of a 1-month-lead ensemble of Artificial Neural Networks (EANN) based on low-frequency climate oscillation indices. The predictand is the February-April (FMA) rainfall in the Brazilian state of Ceará, which is a prominent subject in climate forecasting studies due to its high seasonal predictability. Additionally, the study proposes combining the EANN with dynamical models into a hybrid multi-model ensemble (MME). The forecast verification is carried out through a leave-one-out cross-validation based on 40 years of data. The EANN forecasting skill is compared with traditional statistical models and the dynamical models that compose Ceará’s operational seasonal forecasting system. A spatial comparison showed that the EANN was among the models with the smallest Root Mean Squared Error (RMSE) and Ranked Probability Score (RPS) in most regions. Moreover, the analysis of the area-aggregated reliability showed that the EANN is better calibrated than the individual dynamical models and has better resolution than Multinomial Logistic Regression for above-normal (AN) and below-normal (BN) categories. It is also shown that combining the EANN and dynamical models into a hybrid MME reduces the overconfidence of the extreme categories observed in a dynamically-based MME, improving the reliability of the forecasting system.

54 ENVIRONMENTAL SCIENCES↗

Experimental sloshing pressure impacts in ensemble domain: Transient and stationary statistical characteristics

The present paper focuses on the analysis of impact pressure registrations from repeated model scale sloshing experiments under harmonic rotational excitation. A series of more than 100 experiments, each one encompassing more than 100 impact events, has been conducted seeking the highest feasible repeatability. Different excitation periods, that cover the main features of the impact dynamics, have been considered in a preliminary screening, describing the main features of the impact dynamics. Since, even under a nominally deterministic excitation, the pressure at each impact is characterized by a high variability, a statistical approach is used treating the impact pressure as a stochastic process. For one selected excitation period, the statistical analysis focuses on the ensemble distribution of the maximum pressure during each impact event. Particular attention is given to the evolution of such distributions, in order to detect the variations in the statistical characteristics of the process. This is achieved by, first, identifying the presence and the length of the transient phase and, second, by characterizing the process at stationary state. The statistics of impact pressure for different peaks are discussed mostly in the ensemble domain. Linking the latter with the time domain analysis is made by checking that the problem can be considered “practically ergodic.” The “practical ergodicity” of the process is dealt with by checking to what extent steady state ensemble statistical information can be obtained from a single long run experiment. Statistical checks for correlation and independence of maximum impact pressures are also carried out to test the hypothesis of independent identically distributed random variables. The method of analysis presented in this paper through the considered example case is general in nature and is considered to be highly portable. In particular, it is considered to allow for a more thorough understanding of non-deterministic events such as those considered herein, by looking at them from a sound statistical perspective. The thorough description of the whole experimental setup makes the presented data suitable for comparison purposes and for validation of theoretical/numerical approaches.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Mass assembly of stellar systems and their evolution with the SMA (MASSES). multiplicity and the physical environment in L1448N

We present continuum and molecular line observations at 230 and 345 GHz from the Submillimeter Array (SMA) toward three protostars in the Perseus L1448N region. The data are from the large project “Mass Assembly of Stellar Systems and their Evolution with the SMA.” Three dust continuum sources, Source B, Source NW, and Source A, are detected at both frequencies. These sources have corresponding emission peaks in C{sup 18}O (J=2→1), {sup 13}CO (J=2→1), and HCO{sup +} (J=4→3), and have offsets with N{sub 2}D{sup +} (J=3→2) peaks. High angular resolution data from a complementary continuum survey with the Karl G. Jansky Very Large Array show that Source B is associated with three 8 mm continuum objects, Source NW with two, and Source A remains single. These results suggest that multiplicity in L1448N exists at different spatial scales from a few thousand AU to <100 AU. Velocity gradients in each source obtained from two-dimensional fits to the SMA C{sup 18}O emission are found to be perpendicular to within 20° of the outflow directions as revealed by {sup 12}CO (J=2→1). We have observed that Sources B and NW with multiplicity have higher densities than Source A without multiplicity. This suggests that thermal Jeans fragmentation can be relevant in the fragmentation process. However, we have not observed a difference in the ratio between rotational and gravitational energy between sources with and without multiplicity. We also have not observed a trend between non-thermal velocity dispersions and the level of fragmentation. Our study has provided the first direct and comprehensive comparison between multiplicity and core properties in low-mass protostars, although based on small number statistics.

79 ASTRONOMY AND ASTROPHYSICS↗

THE KINEMATICS OF THE LOCAL GROUP IN A COSMOLOGICAL CONTEXT

Recent observations constrained the tangential velocity of M31 with respect to the Milky Way to be v{sub M31,tan} < 34.4 km s{sup -1}and the radial velocity to be in the range v{sub M31,rad} = -109 {+-} 4.4 km s{sup -1}. In this study we use a large volume high-resolution N-body cosmological simulation (Bolshoi) together with three constrained simulations to statistically study this kinematics in the context of the {Lambda} cold dark matter ({Lambda}CDM). The comparison of the ensembles of simulated pairs with the observed Local Group (LG) at the 1{sigma} level in the uncertainties has been done with respect to the radial and tangential velocities, the reduced orbital energy (e{sub tot}), angular momentum (l{sub orb}), and the dimensionless spin parameter, {lambda}. Our main results are (1) the preferred radial and tangential velocities for pairs in {Lambda}CDM are v{sub r} = -80 {+-} 20 km s{sup -1} and v{sub t} = 50 {+-} 10 km s{sup -1}, (2) pairs around that region are 3-13 times more common than pairs within the observational values, (3) 15%-24% of LG-like pairs in {Lambda}CDM have energy and angular momentum consistent with observations, while (4) 9%-13% of pairs in the same sample show similar values in the inferred dimensionless spin parameter. It follows that within current observational uncertainties the quasi-conserved quantities that characterize the orbit of the LG, i.e., e{sub tot}, l{sub orb}, and {lambda}, do not challenge the standard {Lambda}CDM model, but the model is in tension with regard to the actual values of the radial and tangential velocities. This might hint to a problem of the {Lambda}CDM model to reproduce the observed LG.

79 ASTRONOMY AND ASTROPHYSICS↗

Noise-resolution tradeoffs in x-ray CT imaging: A comparison of penalized alternating minimization and filtered backprojection algorithms

Purpose: In comparison with conventional filtered backprojection (FBP) algorithms for x-ray computed tomography (CT) image reconstruction, statistical algorithms directly incorporate the random nature of the data and do not assume CT data are linear, noiseless functions of the attenuation line integral. Thus, it has been hypothesized that statistical image reconstruction may support a more favorable tradeoff than FBP between image noise and spatial resolution in dose-limited applications. The purpose of this study is to evaluate the noise-resolution tradeoff for the alternating minimization (AM) algorithm regularized using a nonquadratic penalty function. Methods: Idealized monoenergetic CT projection data with Poisson noise were simulated for two phantoms with inserts of varying contrast (7%-238%) and distance from the field-of-view (FOV) center (2-6.5 cm). Images were reconstructed for the simulated projection data by the FBP algorithm and two penalty function parameter values of the penalized AM algorithm. Each algorithm was run with a range of smoothing strengths to allow quantification of the noise-resolution tradeoff curve. Image noise is quantified as the standard deviation in the water background around each contrast insert. Modulation transfer functions (MTFs) were calculated from six-parameter model fits to oversampled edge-spread functions defined by the circular contrast-insert edges as a metric of local resolution. The integral of the MTF up to 0.5 lp/mm was adopted as a single-parameter measure of local spatial resolution. Results: The penalized AM algorithm noise-resolution tradeoff curve was always more favorable than that of the FBP algorithm. While resolution and noise are found to vary as a function of distance from the FOV center differently for the two algorithms, the ratio of noises when matching the resolution metric is relatively uniform over the image. The ratio of AM-to-FBP image variances, a predictor of dose-reduction potential, was strongly dependent on the shape of the AM's nonquadratic penalty function and was also strongly influenced by the contrast of the insert for which resolution is quantified. Dose-reduction potential, reported here as the fraction (%) of FBP dose necessary for AM to reconstruct an image with comparable noise and resolution, for one penalty parameter value of the AM algorithm was found to vary from 70% to 50% for low-contrast and high-contrast structures, respectively, and from 70% to 10% for the second AM penalty parameter value. However, the second penalty, AM-700, was found to suffer from poor low-contrast resolution when matching the high-contrast resolution metric with FBP. Conclusions: The results of this simulation study imply that penalized AM has the potential to reconstruct images with similar noise and resolution using a fraction (10%-70%) of the FBP dose. However, this dose-reduction potential depends strongly on the AM penalty parameter and the contrast magnitude of the structures of interest. In addition, the authors' results imply that the advantage of AM can be maximized by optimizing the nonquadratic penalty function to the specific imaging task of interest. Future work will extend the methods used here to quantify noise and resolution in images reconstructed from real CT data.

60 APPLIED LIFE SCIENCES↗

Mapathons versus automated feature extraction: a comparative analysis for strengthening immunization microplanning

Background: Social instability and logistical factors like the displacement of vulnerable populations, the difficulty of accessing these populations, and the lack of geographic information for hard-to-reach areas continue to serve as barriers to global essential immunizations (EI). Microplanning, a population-based, healthcare intervention planning method has begun to leverage geographic information system (GIS) technology and geospatial methods to improve the remote identification and mapping of vulnerable populations to ensure inclusion in outreach and immunization services, when feasible. We compare two methods of accomplishing a remote inventory of building locations to assess their accuracy and similarity to currently employed microplan line-lists in the study area. Methods: The outputs of a crowd-sourced digitization effort, or mapathon, were compared to those of a machine-learning algorithm for digitization, referred to as automatic feature extraction (AFE). The following accuracy assessments were employed to determine the performance of each feature generation method: (1) an agreement analysis of the two methods assessed the occurrence of matches across the two outputs, where agreements were labeled as “befriended” and disagreements as “lonely”; (2) true and false positive percentages of each method were calculated in comparison to satellite imagery; (3) counts of features generated from both the mapathon and AFE were statistically compared to the number of features listed in the microplan line-list for the study area; and (4) population estimates for both feature generation method were determined for every structure identified assuming a total of three households per compound, with each household averaging two adults and 5 children. Results: The mapathon and AFE outputs detected 92,713 and 53,150 features, respectively. A higher proportion (30%) of AFE features were befriended compared with befriended mapathon points (28%). The AFE had a higher true positive rate (90.5%) of identifying structures than the mapathon (84.5%). The difference in the average number of features identified per area between the microplan and mapathon points was larger (t = 3.56) than the microplan and AFE (t = -2.09) (alpha = 0.05). Conclusions: Our findings indicate AFE outputs had higher agreement (i.e., befriended), slightly higher likelihood of correctly identifying a structure, and were more similar to the local microplan line-lists than the mapathon outputs. These findings suggest AFE may be more accurate for identifying structures in high-resolution satellite imagery than mapathons. However, they both had their advantages and the ideal method would utilize both methods in tandem.

59 BASIC BIOLOGICAL SCIENCES↗