Search NASA⌕ Search

SEARCH · Search NASA

Results for “LDA”

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

AI-powered topic modeling: comparing LDA and BERTopic in analyzing opioid-related cardiovascular risks in women

Topic modeling is a crucial technique in natural language processing (NLP), enabling the extraction of latent themes from large text corpora. Traditional topic modeling, such as Latent Dirichlet Allocation (LDA), faces limitations in capturing the semantic relationships in the text document although it has been widely applied in text mining. BERTopic, created in 2022, leveraged advances in deep learning and can capture the contextual relationships between words. In this work, we integrated Artificial Intelligence (AI) modules to LDA and BERTopic and provided a comprehensive comparison on the analysis of prescription opioid-related cardiovascular risks in women. Opioid use can increase the risk of cardiovascular problems in women such as arrhythmia, hypotension etc. 1,837 abstracts were retrieved and downloaded from PubMed as of April 2024 using three Medical Subject Headings (MeSH) words: “opioid,” “cardiovascular,” and “women.” Machine Learning of Language Toolkit (MALLET) was employed for the implementation of LDA. BioBERT was used for document embedding in BERTopic. Eighteen was selected as the optimal topic number for MALLET and 23 for BERTopic. ChatGPT-4-Turbo was integrated to interpret and compare the results. The short descriptions created by ChatGPT for each topic from LDA and BERTopic were highly correlated, and the performance accuracies of LDA and BERTopic were similar as determined by expert manual reviews of the abstracts grouped by their predominant topics. The results of the t-SNE (t-distributed Stochastic Neighbor Embedding) plots showed that the clusters created from BERTopic were more compact and well-separated, representing improved coherence and distinctiveness between the topics. Our findings indicated that AI algorithms could augment both traditional and contemporary topic modeling techniques. In addition, BERTopic has the connection port for ChatGPT-4-Turbo or other large language models in its algorithm for automatic interpretation, while with LDA interpretation must be manually, and needs special procedures for data pre-processing and stop words exclusion. Therefore, while LDA remains valuable for large-scale text analysis with resource constraints, AI-assisted BERTopic offers significant advantages in providing the enhanced interpretability and the improved semantic coherence for extracting valuable insights from textual data.

Research & Experimental Medicine↗

Fast Solution in Sparse LDA for Binary Classification

An algorithm that performs sparse linear discriminant analysis (Sparse-LDA) finds near-optimal solutions in far less time than the prior art when specialized to binary classification (of 2 classes). Sparse-LDA is a type of feature- or variable- selection problem with numerous applications in statistics, machine learning, computer vision, computational finance, operations research, and bio-informatics. Because of its combinatorial nature, feature- or variable-selection problems are NP-hard or computationally intractable in cases involving more than 30 variables or features. Therefore, one typically seeks approximate solutions by means of greedy search algorithms. The prior Sparse-LDA algorithm was a greedy algorithm that considered the best variable or feature to add/ delete to/ from its subsets in order to maximally discriminate between multiple classes of data. The present algorithm is designed for the special but prevalent case of 2-class or binary classification (e.g. 1 vs. 0, functioning vs. malfunctioning, or change versus no change). The present algorithm provides near-optimal solutions on large real-world datasets having hundreds or even thousands of variables or features (e.g. selecting the fewest wavelength bands in a hyperspectral sensor to do terrain classification) and does so in typical computation times of minutes as compared to days or weeks as taken by the prior art. Sparse LDA requires solving generalized eigenvalue problems for a large number of variable subsets (represented by the submatrices of the input within-class and between-class covariance matrices). In the general (fullrank) case, the amount of computation scales at least cubically with the number of variables and thus the size of the problems that can be solved is limited accordingly. However, in binary classification, the principal eigenvalues can be found using a special analytic formula, without resorting to costly iterative techniques. The present algorithm exploits this analytic form along with the inherent sequential nature of greedy search itself. Together this enables the use of highly-efficient partitioned-matrix-inverse techniques that result in large speedups of computation in both the forward-selection and backward-elimination stages of greedy algorithms in general.

Moghaddam, Baback↗

High-resolution LDA measurements of Reynolds stress in boundary layers and wakes

The turbulent character of the boundary layer and wake associated with an airfoil has been studied at a Reynolds number of 1,000,000 and a Mach number of 0.1. To accomplish these measurements, a unique laser Doppler anemometer (LDA) has been developed that is capable of sensing two velocity components from a remote distance of 2.13 m. Using special simultaneity logic and counter-type signal processors, the geometrical features of the LDA have been exploited to provide variable spatial resolution as low as 0.2 mm. By combining the LDA with an on-line computerized data acquisition and display system, it has been possible to measure mean velocity and Reynolds stress tensor distribution at several locations along the upper surface of a 0.9-m-chord, flapped airfoil installed in the Ames 7- by 10-Foot Wind Tunnel.

Orloff, K. L.↗

LDA measurements in a Mach 2 flow over a rearward facing step with staged transverse injection

Measurements of the mean velocity field and selected turbulence statistics have been obtained via 2D LDA in a Mach 2 flow over a rearward facing step with downstream transverse injection. Axial mean velocity profiles, profiles of the axial and normal RMS levels, and the correlation coefficient between the axial and normal fluctuating components are presented for locations upstream of the step. These data, through comparison with other reported measurements in zero pressure gradient compressible boundary layers indicate that the boundary layers upstream of the step are consistent with 'quasi-equilibrium' turbulent boundary layers with the adiabatic wall boundary condition. Mean velocity field measurements on the symmetry plane of the tunnel are compared to laser induced iodine fluorescence measurements reported in the literature and obtained in the same facility. This comparison demonstrates the quality of the present LDA data set and shows that particle lag is not significant in the LDA measurements despite the complex nature of the downstream flow.

Wang, J.-A.↗

High-Resolution LDA Measurements of Reynolds Stress in Boundary Layers and Wakes

The turbulent character of the boundary layer and wake associated with an airfoil has been studied at a Reynolds number of 10(exp 6) and a Mach number of 0.1. To accomplish these measurements, a unique laser Doppler anemometer (LDA) has been developed that is capable of sensing two velocity components from a remote distance of 2.13 m. Using special simultaneity logic and counter-type signal processors, the geometrical features of the LDA have been exploited to provide variable spatial resolution as low as 0.2 mm. By combining the LDA with an on-line computerized data acquisition and display system, it has been possible to measure mean velocity and Reynolds stress tensor distribution at several locations along the upper surface of a 0.9 m chord, flapped airfoil installed in the Ames 7- by 10-Foot Wind Tunnel.

Orloff, Kenneth↗

Three component velocity measurements in an axisymmetric jet using LDA

A commercial three-component laser Doppler anemometer (LDA) has been used to acquire a detailed set of three-dimensional mean and fluctuating velocity measurements in a low-speed air jet entering a stagnant ambient, over the first 15 jet exit diameters along the jet trajectory. These data are consistent with previous measurements in axisymmetric, turbulent jets. Mean velocity and Reynolds stress data approach a self-preserving behavior by x/D equal to about 15. However, the RMS turbulence fluctuations were not self-preserving at this axial location, as expected based upon previous experimental studies. These data confirm the ability to obtain reliable three-dimensional velocity data using the present three-dimensional LDA system.

Kuhlman, John M.↗

An LDA investigation of three-dimensional normal shock-boundary layer interactions in a corner

Nonintrusive, three-dimensional, measurements have been made of a normal shock wave-turbulent boundary layer interaction. The measurements were made in the corner of the test section of a continuous supersonic wind tunnel in which a normal shock wave had been stabilized. LDA, surface pressure measurement and flow visualization techniques were employed for two freestream Mach number test cases: 1.6 and 1.3. The former contained separated flow regions and a system of shock waves. The latter was found to be far less complicated. The reported results are believed to accurately define the flow physics of each case and may be used as benchmark data to verify three-dimensional computer codes.

Chriss, R. M.↗

LDA measurement of the passage flow field in an annular airfoil cascade

Models to predict the complex three-dimensional flow through turbomachine blade rows are being developed. To verify these models and direct necessary refinements, it is necessary that predictions be correlated with data obtained in experiments which model the fundamental three-dimensional blade row flow phenomena. This paper describes a series of experiments performed in a large scale, subsonic, annular cascade facility specifically designed to provide such data. In particular, the effect of incidence angle on the three-dimensional passage flow field through an annular cascade of cambered airfoils is investigated and quantified, accomplished by obtaining detailed and expensive LDA data. These data demonstrate and quantify the development of the passage vortices through the airfoil passage and their strong interaction with the endwall boundary layers.

Stauter, R. C.↗

Opposed jet diffusion flames of nitrogen-diluted hydrogen vs air - Axial LDA and CARS surveys; fuel/air rates at extinction

An experimental study of H-air counterflow diffusion flames (CFDFs) is reported. Coaxial tubular opposed jet burners were used to form dish-shaped CFDFs centered by opposing laminar jets of H2/N2 and air in an argon bath at 1 atm. Jet velocities for extinction and flame restoration limits are shown versus input H2 concentration. LDA velocity data and CARS temperature and absolute N2, O2 density data give detailed flame structure on the air side of the stagnation point. The results show that air jet velocity is a more fundamental and appropriate measure of H2-air CFDF extinction than input H2 mass flux or fuel jet velocity. It is proposed that the observed constancy of air jet velocity for fuel mixtures containing 80 to 100 percent H2 measure a maximum, kinetically controlled rate at which the CFDF can consume oxygen in air. Fuel velocity mainly measures the input jet momentum required to center an H2/N2 versus air CFDF.

Pellett, G. L.↗

An LDA investigation of the normal shock wave boundary layer interaction

Nonintrusive measurements have been made of two normal shock wave-boundary layer interactions. Two-dimensional measurements were made throughout the interaction region while three-dimensional measurements were made in the vicinity of the shock wave. The measurements were made in the corner of the test section of a continuous flow supersonic wind tunnel in which a normal shock wave had been stabilized. LDA, surface pressure measurement and flow visualization techniques were employed for two freestream Mach number test cases: 1.6 and 1.3. The former contained separated flow regions and a system of shock waves. The latter was found to be far less complicated. The reported results define the flowfield structure in detail for each case.

Chriss, R. M.↗

An experimental technique for performing 3-D LDA measurements inside whirling annular seals

During the last several years, the Fluid Mechanics Division of the Turbomachinery Laboratory at Texas A&M University has developed a rather unique facility with the experimental capability for measuring the flow field inside journal bearings, labyrinth seals, and annular seals. The facility consists of a specially designed 3-D LDA system which is capable of measuring the instantaneous velocity vector within 0.2 mm of a wall while the laser beams are aligned almost perpendicular to the wall. This capability was required to measure the flow field inside journal bearings, labyrinth seals, and annular seals. A detailed description of this facility along with some representative results obtained for a whirling annular seal are presented.

Morrison, Gerald L.↗

Effect of the electronic charge gap on LO bond-stretching phonons in undoped La 2 CuO 4 calculated using LDA + U

Typical density-functional theory calculations that wrongly predict undoped cuprates to be metallic also predict Cu-O half- and full-breathing phonon energies that are significantly softer than observed, presumably because of weak on-site Coulomb repulsion on the Cu 3d orbitals. We used DFT+U calculations with antiferromagnetic supercells of La 2 CuO 4 to establish correlation between the on-site repulsion strength, tuned via adjusting the value of U, and phonon dispersions. Here, we find that breathing and half-breathing phonons reach experimental values when U is tuned to obtain the correct optical gap and magnetic moments. We demonstrate that using distorted supercells within DFT+U is a promising framework to model phonons in undoped cuprates and other perovskite oxides with complex, interrelated structural and electronic degrees of freedom.

36 MATERIALS SCIENCE↗

LDA seeding system for the Langley Low Turbulence Pressure Tunnel

A Laser Velocimetry (LV) seeding system was specifically developed for the Langley Low Turbulence Wind Tunnel (LTPT), and it has been successfully used for LV measurements in two major tests (Juncture Flow Experiment and Gortler Experiment). The LTPT is capable of operating at Mach numbers from 0.05 to 0.50 and unit Reynolds numbers from 100,000 to 15,000,000 per foot. The test section is 3 feet wide and 7.5 feet high. The turbulence level in the test section is relatively low because of the high contraction ratio and because of the nine turbulence reduction screens in the settling chamber. A primary requirement of the seeding system was that the seeding material not contaminate or damage in any way these screens. Both solid and liquid seeding systems were evaluated, and the results are presented. They can provide some guidelines for setting up seeding systems in other similar tunnels.

Scheiman, J.↗

LDA measurement of the passage flow field in a 3-D airfoil cascade

Three-dimensional internal flow computational models are currently being developed to predict the flow through turbomachinery blade rows. For these codes to be of quantitative value, they must be verified with data obtained in experiments which model the fundamental flow phenomena. In this paper, the complete three-dimensional flow field through a subsonic annular cascade of cambered airfoils is experimentally quantified. In particular, detailed three-dimensional data are obtained to quantify the inlet velocity profile, the cascade passage velocity field, and the exit region flow field. The primary instrumentation for acquiring these data is a single-channel Laser Doppler Anemometer operating in the backscatter mode, with chordwise distributions of airfoil surface static pressure taps also utilized. Appropriate data are correlated with predictions from the MERIDL/TSONIC codes.

Stauter, R. C.↗

3-D LDA study of a rectangular jet

The flow field of a rectangular jet with a 2:1 aspect ratio was studied at an axial Reynolds number of 100,000 (Mach number 0.09) using three-dimensional laser Doppler velocimetry. The flow field survey resulted in mean velocity vector field plots and contour plots of the Reynolds stress tensor components. This paper presents contour plots in the planes of the jet minor and major axes at different axial locations. These data contribute substantially to currently available data of jet flow fields and will provide a valuable database for three-dimensional modeling.

Morrison, Gerald L.↗

An LDA (Laser-Doppler Anemometry) investigation of three-dimensional normal shock wave boundary-layer interactions

Nonintrusive measurements were made of a normal shock wave/boundary layer interaction. Two dimensional measurements were made throughout the interaction region while 3-D measurements were made in the vicinity of the shock wave. The measurements were made in the corner of the test section of a continuous supersonic wind tunnel in which a normal shock wave had been stabilized. Laser Doppler Anemometry, surface pressure measurement and flow visualization techniques were employed for two freestream Mach number test cases: 1.6 and 1.3. The former contained separated flow regions and a system of shock waves. The latter was found to be far less complicated. The results define the flow field structure in detail for each case.

Chriss, R. M.↗

Vorticity, turbulence production, and turbulence induced accelerations in a rectangular jet as measured using 3-D LDA

The flow field of a rectangular jet with a 4:1 aspect ratio (50.4 x 12.7 mm) was studied at a Reynolds number of 100,000 (Mach number 0.09) using a 3-D laser Doppler anemometer system. Measurements were performed along the major and minor axis planes and at various downstream cross-sections of the jet. The mean velocity vector and entire Reynolds stress tensor were measured and presented in a previous publication. The present work presents the vorticity vector, turbulence production, and turbulence induced acceleration vector distributions which were calculated from the previously presented data.

Morrison, Gerald L.↗