Search NASA⌕ Search

SEARCH · Search NASA

Results for “feature selection”

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 721 records · Page 40

Kinetic Control via Binding Sites within the Confined Space of Metal Metalloporphyrin‐Frameworks for Enhanced Shape‐Selectivity Catalysis

Abstract One striking feature of enzyme is its controllable ability to trap substrates via synergistic or cooperative binding in the enzymatic pocket, which renders the shape‐selectivity of product by the confined spatial environment. The success of shape‐selective catalysis relies on the ability of enzyme to tune the thermodynamics and kinetics for chemical reactions. In emulation of enzyme's ability, we showcase herein a targeting strategy with the substrate being anchored on the internal pore wall of metal‐organic frameworks (MOFs), taking full advantage of the sterically kinetic control to achieve shape‐selectivity for the reactions. For this purpose, a series of binding site‐accessible metal metalloporphyrin‐frameworks (MMPFs) have been investigated to shed light on the nature of enzyme‐mimic catalysis. They exhibit a different density of binding sites that are well arranged into the nanospace with corresponding distances of opposite binding sites. Such a structural specificity results in a facile switch in selectivity from an exclusive formation of the thermodynamically stable product to the kinetic product. Thus, the proposed targeting strategy, based on the combination of porous materials and binding events, paves a new way to develop highly efficient heterogeneous catalysts for shifting selectivity.

Zhang, Weijie↗

H{alpha} AND 4000 A BREAK MEASUREMENTS FOR {approx}3500 K-SELECTED GALAXIES AT 0.5 < z < 2.0

We measure spectral features of {approx}3500 K-selected galaxies at 0.5 < z < 2.0 from high-quality medium-band photometry using a new technique. First, we divide the galaxy sample in 32 subsamples based on the similarities between the full spectral energy distributions (SEDs) of the galaxies. For each of these 32 galaxy types we construct a composite SED by de-redshifting and scaling the observed photometry. This approach increases the signal-to-noise ratio and sampling of galaxy SEDs and allows for model-independent stellar population studies. The composite SEDs are of spectroscopic quality and facilitate-for the first time-H{alpha} measurements for a large magnitude-limited sample of distant galaxies. The line widths indicate a photometric redshift uncertainty of {Delta}z < 0.02 Multiplication-Sign (1 + z). The composite SEDs also show the Balmer and 4000 A breaks, Mg II absorption at {approx}2800 A, the dust absorption feature at 2175 A, and blended [O III]+H{beta} emission. We compare the total equivalent width of H{alpha}, [N II], and [S II] (W{sub H{alpha}+}) with the strength of the 4000 A break (D(4000)) and the best-fit specific star formation rate, and find that all these properties are strongly correlated. This is a reassuring result, as currently most distant stellar population studies are based on just continuum emission. Furthermore, the relation between W{sub H{alpha}+} and D(4000) provides interesting clues to the star formation histories of galaxies, as these features are sensitive to different stellar ages. We find that the correlation between W{sub H{alpha}+} and D(4000) at 0.5 < z < 2.0 is similar to z {approx} 0 and that the suppression of star formation in galaxies at z < 2 is generally not abrupt, but a gradual process.

79 ASTRONOMY AND ASTROPHYSICS↗

Study of Turbofan Engines Designed for Low Enery Consumption

Subsonic transport turbofan engine design and technology features which have promise of improving aircraft energy consumption are described. Task I addressed the selection and evaluation of features for the CF6 family of engines in current aircraft, and growth models of these aircraft. Task II involved cycle studies and the evaluation of technology features for advanced technology turbofans, consistent with initial service in 1985. Task III pursued the refined analysis of a specific design of an advanced technology turbofan engine selected as the result of Task II studies. In all of the above, the impact upon aircraft economics, as well as energy consumption, was evaluated. Task IV summarized recommendations for technology developments which would be necessary to achieve the improvements in energy consumption identified.

Neitzel, R. E.↗

Teacher Enhancement Institute

During the 1980's, a period of intense concern over educational quality in the United States, few indicators of U.S. student achievement garnered the interest of policy makers and pundits as successfully as the results of international testing in mathematics and science. This concern was so great that as a part of the Goals 2000 initiative, President George Bush indicated that 'By the year 2000, U.S. students should be first in the world in mathematics and science.' The Clinton Administration is placing a major emphasis, not only on rigorous academic standards and creating a new system for assessing students' progress, but also including professional development as a major focus. The argument being that teachers need more sustained, intensive training to prepare them to teach to higher standards. Executive order 12821 mandates that national laboratories 'assist in the mathematics and science education of our Nation's students, teachers, parents and the public by establishing programs at their agency to provide for training elementary and secondary school teachers to improve their knowledge of mathematics and science'. These and other issues led to the development of ideas for a project that addresses the need for excellence in mathematics, science and technology instruction. In response to these initiatives the NASA/LaRC Teacher Enhancement Institute was proposed. The TEI incorporated systemic reform perspectives, enhanced content knowledge for teachers, and teacher preparation. Emphasis was also placed on recruiting those educators who teach in impoverished urban school districts with at-risk student populations who have been traditionally under represented in science, mathematics, technology and engineering. Participants in the Teacher Enhancement Institute were 37 teachers from grades K-8, teaching in Region 2 in the state of Virginia, as well as 2 preservice teachers from Norfolk State University and one teacher from Dublin, Virginia, where a Science/Mathematics model school has been established. Teachers selected for this project represented school systems where income levels are extremely low, and students served tend not to receive innovative instruction in mathematics and science and their use of technology is limited. The Teacher Enhancement Institute contained several features, that when combined, allowed for a unique experience. Some of these features included local teachers, administrators and school board members as presenters, instruction and use of technology every day, tours of select features of the research facility, briefings by NASA/LaRC scientists, engineers and researchers as well as individuals from the Continuous Electron Beam Accelerator Facility (CEBAF). Another unique feature of this program is to have participants convene on three separate occasions throughout the academic year to discuss strategies for information dissemination and implementation results. Teachers' attitudes towards the use of technology, their ability to develop lessons using technology and their ability to develop lessons using information obtained through TEI were assessed using instruments developed by TEI summer faculty members. Data from these instruments were analyzed and reported in a final report submitted to the director of the Office of Education.

Marshall-Bradley, Tina↗

Highly Selective Adsorption of Ethylene over Ethane in a MOF Featuring the Combination of Open Metal Site and -Complexation

The introduction of the combination of open metal site (OMS) and -complexation into MOF has led to very high ethylene/ethane adsorption selectivity at 318K, as illustrated in the context of MIL-101-Cr-SO 3 Ag. The interactions with ethylene from both OMS and -complexation in MIL-101-Cr-SO 3 Ag have been investigated by in situ IR spectroscopic studies and computational calculations, which suggest -complexation contributes dominantly to the high ethylene/ethane adsorption selectivity.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Multilayer Microfluidic Devices Created From A Single Photomask

The time and expense associated with high quality photomask production can discourage the creation of multilayer microfluidic devices, as each layer currently requires a separate photomask. Here we describe an approach in which multilayer microfabricated devices can be created from a single photomask. The separate layers and their corresponding alignment marks are arranged in separate halves of the mask for two layer devices or quadrants for four layer devices. Selective exposure of the photomask features and rotation of the device substrate between exposures result in multiple copies of the devices on each wafer. Subsequent layers are aligned to patterned features on the substrate with the same alignment accuracy as when multiple photomasks are used. We demonstrate this approach for fabricating devices employing multilayer soft lithography (MSL) for pneumatic valving. MSL devices containing as many as 5 layers (4 aligned fluidic layers plus a manually aligned control layer) were successfully created using this approach. Device design is also modularized, enabling the presence or absence of features as well as channel heights to be selected independently from one another. The use of a single photomask to create multilayer devices results in a dramatic savings of time and/or money required to advance from device design to completed prototype.

47 OTHER INSTRUMENTATION↗

Characterization of ultrafast laser-ablation plasma plumes at various Ar ambient pressures

Expansion dynamics and internal plume structures of fs laser ablated brass plasma in Ar at various pressure levels ranging from vacuum to atmospheric were studied using multitude of diagnostic tools including time resolved and time integrated 2-dimensional imaging, optical time of flight measurements and visible emission spectroscopy. Temporal evolution of excited Cu and Zn species in the plume were imaged using band pass interference filters and compared its hydrodynamic expansion features with spectrally integrated images of the plume. 2D imaging coupled with monochromatic line selection showed several interesting features at various pressure levels which include velocity differences among the plume species, emission intensity distribution, plasma temperature, electron density etc. Plume confinement, enhanced signal intensity, and dual peak structures in time-of-flight profiles were observed at intermediate pressure range of ~10 Torr. Optimum signal to background ratio was also observed in this pressure range. As a result, possible mechanisms for observed changes in plume shape, optical emission intensity and dual peak structures in time-of-flight profiles were discussed.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

System Would Detect Foreign-Object Damage in Turbofan Engine

A proposed data-fusion system, to be implemented mostly in software, would further process the digitized and preprocessed outputs of sensors in a turbofan engine to detect foreign-object damage (FOD) [more precisely, damage caused by impingement of such foreign objects as birds, pieces of ice, and runway debris]. The proposed system could help a flight crew to decide what, if any, response is necessary to complete a flight safely, and could aid mechanics in deciding what post-flight maintenance action might be needed. The sensory information to be utilized by the proposed system would consist of (1) the output of an accelerometer in an engine-vibration-monitoring subsystem and (2) features extracted from a gas path analysis. ["Gas path analysis" (GPA) is a term of art that denotes comprehensive analysis of engine performance derived from readings of fuel-flow meters, shaft-speed sensors, temperature sensors, and the like.] The acceleration signal would first be processed by a wavelet-transform-based algorithm, using a wavelet created for the specific purpose of finding abrupt FOD-induced changes in noisy accelerometer signals. Two additional features extracted would be the amplitude of vibration (determined via a single- frequency Fourier transform calculated at the rotational speed of the engine), and the rate of change in amplitude due to an FOD-induced rotor imbalance. This system would utilize two GPA features: the fan efficiency and the rate of change of fan efficiency with time. The selected GPA and vibrational features would be assessed by two fuzzy-logic inference engines, denoted the "Gas Path Expert" and the "Vibration Expert," respectively (see Figure 1). Each of these inference engines would generate a "possibility" distribution for occurrence of an FOD event: Each inference engine would assign, to its input information, degrees of membership, which would subsequently be transformed into basic probability assignments for the gas path and vibration components. The outputs of the inference engines would be fused by use of Dempster s combination algorithm (more precisely, an algorithm, based on the Dempster-Shafer-Yager theory of evidence, for fusing uncertain or imprecise information) to provide a reduced body of information to a human or computer decision maker. Figure 2 depicts some outputs generated in response to simulated accelerometer and GPA signals

Torso, James A.↗

Dynamic logistic regression and variable selection: Forecasting and contextualizing civil unrest

Civil unrest can range from peaceful protest to violent furor, and researchers are working to monitor, forecast, and assess such events to allocate resources better. Twitter has become a real-time data source for forecasting civil unrest because millions of people use the platform as a social outlet. Additionally, daily word counts are used as model features, and predictive terms contextualize the reasons for the protest. To forecast civil unrest and infer the reasons for the protest, we consider the problem of Bayesian variable selection for the dynamic logistic regression model and propose using penalized credible regions to select parameters of the updated state vector. This method avoids the need for shrinkage priors, is scalable to high-dimensional dynamic data, and allows the importance of variables to vary in time as new information becomes available. A substantial improvement in both precision and F1-score using this approach is demonstrated through simulation. Finally, we apply the proposed model fitting and variable selection methodology to the problem of forecasting civil unrest in Latin America. Our dynamic logistic regression approach shows improved accuracy compared to the static approach currently used in event prediction and feature selection.

97 MATHEMATICS AND COMPUTING↗

On the use of graph search techniques for the analysis of extreme-scale combustion simulation data

With the continuous increase in high performance computing capabilities, simulations are becoming ever larger and more complex, using bigger domains, tracking more variables, and producing more time steps. This increase in the ranges of spatial and temporal simulation scales results in data that presents significant challenges to as well as new opportunities for the visualization and data analysis community. For example, highly-localized, intermittent events (such as the formation of ignition kernels in turbulent combustion) may be caused by interactions between multiple variables across a series of time steps, making both their definition and their extraction difficult, particularly at scale. This paper introduces an intuitive framework to support the identification, characterization, and tracking of such complex, multivariate, temporally evolving events in large-scale simulations. In a pre-processing step, we use topological techniques to create a hierarchical family of feature definitions for each variable of interest. Subsequently, we select a particular set of features for analysis and, using overlap-based metrics, we generate an attributed relational graph (ARG) capturing the relationships between different variables both within one and across multiple time steps. Finally, we leverage subgraph-isomorphism search heuristics to identify patterns in the ARG that characterize interesting events. We demonstrate the power of this approach by analyzing a large-scale turbulent combustion simulation.

McLendon, W. C.↗

Narrow band wavelength selective filter using grating assisted single ring resonator

This paper illustrates a filter configuration which uses a single ring resonator of larger radius connected to a grating resonator at its drop port to achieve single wavelength selectivity and switching property with spectral features suitable for on-chip wavelength selection applications. The proposed configuration is expected to find applications in silicon photonics devices such as, on-chip external cavity lasers and multi analytic label-free biosensors. The grating resonator has been designed for a high Q-factor, high transmittivity, and minimum loss so that the wavelength selectivity of the device is improved. The proof-of-concept device has been demonstrated on a Silicon-on-Insulator (SOI) platform through electron beam lithography and Reactive Ion Etching (RIE) process. The transmission spectrum shows narrow band single wavelength selection and switching property with a high Free Spectral Range (FSR) ∼60 nm and side band rejection ratio >15 dB.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Selective binding of choline by a phosphate-coordination-based triple helicate featuring an aromatic box

In nature, proteins have evolved sophisticated cavities tailored for capturing target guests selectively among competitors of similar size, shape, and charge. The fundamental principles guiding the molecular recognition, such as self-assembly and complementarity, have inspired the development of biomimetic receptors. In the current work, we report a self-assembled triple anion helicate (host 2) featuring a cavity resembling that of the choline-binding protein ChoX, as revealed by crystal and density functional theory (DFT)-optimized structures, which binds choline in a unique dual-site-binding mode. Here, this similarity in structure leads to a similarly high selectivity of host 2 for choline over its derivatives, as demonstrated by the NMR and fluorescence competition experiments. Furthermore, host 2 is able to act as a fluorescence displacement sensor for discriminating choline, acetylcholine, l-carnitine, and glycine betaine effectively.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Operational plans for life science payloads - From experiment selection through postflight reporting

Key features of operational plans developed in a study of the Space Shuttle era life science payloads program are presented. The data describes the overall acquisition, staging, and integration of payload elements, as well as program implementation methods and mission support requirements. Five configurations were selected as representative payloads: (a) carry-on laboratories - medical emphasis experiments, (b) mini-laboratories - medical/biology experiments, (c) seven-day dedicated laboratories - medical/biology experiments, (d) 30-day dedicated laboratories - Regenerative Life Support Evaluation (RLSE) with selected life science experiments, and (e) Biomedical Experiments Scientific Satellite (BESS) - extended duration primate (Type I) and small vertebrate (Type II) missions. The recommended operational methods described in the paper are compared to the fundamental data which has been developed in the life science Spacelab Mission Simulation (SMS) test series. Areas assessed include crew training, experiment development and integration, testing, data-dissemination, organization interfaces, and principal investigator working relationships.

Mccollum, G. W.↗

Quantifying Community Assembly Processes and Identifying Features that Impose Them

Across a set of ecological communities connected to each other through organismal dispersal (a ‘meta-community’), turnover in composition is governed by (ecological) Drift, Selection, and Dispersal Limitation. Quantitative estimates of these processes remain elusive, but would represent a common currency needed to unify community ecology. Using a novel analytical framework we quantitatively estimate the relative influences of Drift, Selection, and Dispersal Limitation on subsurface, sediment-associated microbial meta-communities. The communities we study are distributed across two geologic formations encompassing ~12,500m3 of uranium-contaminated sediments within the Hanford Site in eastern Washington State. We find that Drift consistently governs ~25% of spatial turnover in community composition; Selection dominates (governing ~60% of turnover) across spatially-structured habitats associated with fine-grained, low permeability sediments; and Dispersal Limitation is most influential (governing ~40% of turnover) across spatially-unstructured habitats associated with coarse-grained, highly-permeable sediments. Quantitative influences of Selection and Dispersal Limitation may therefore be predictable from knowledge of environmental structure. To develop a system-level conceptual model we extend our analytical framework to compare process estimates across formations, characterize measured and unmeasured environmental variables that impose Selection, and identify abiotic features that limit dispersal. Insights gained here suggest that community ecology can benefit from a shift in perspective; the quantitative approach developed here goes beyond the ‘niche vs. neutral’ dichotomy by moving towards a style of natural history in which estimates of Selection, Dispersal Limitation and Drift can be described, mapped and compared across ecological systems.

ecological drift↗

Summary of the 19th International Conference on Arabidopsis Research (July 23-27, 2008 in Montreal, Canada)

The 19th International Conference on Arabidopsis Research was a successful meeting attended by 815 scientists from around the world including 322 from the United States, 146 attendees from Canada, 179 from Europe, 134 from Asia, and 34 from a combination of Australia, South America, Africa and the Middle East. The scientific program was of excellent quality featuring 64 talks, including 41 from invited speakers. The Keynote Lecture, delivered by Chris Somerville (Energy Biosciences Institute/UC Berkeley) was particularly relevant to US agriculture and energy research and was titled The Development of Cellulosic Biofuels. There were also 6 community-organized workshops featuring 30 additional talks on topics including Frontiers in Plant Systems Biology, Sources and strategies for Gene Structure, Gene Function, and Metabolic Pathway annotation at TAIR and AraCyc, Advanced Bioinformatic Resources for Arabidopsis, Laser Microtechniques and Applications with Arabidopsis, Plant Proteomics- Tools, Approaches, Standards and Breakthroughs in Studying the Proteome, and Phytohormone Biosynthesis and Signal Transduction. Conference organizers arranged a special seminar by Jim Collins (head of the Directorate of Biosciences at NSF) to provide a community discussion forum regarding the future of Arabidopsis research. Approximately 575 posters were presented in topic areas including, among others, Development, Signal Transduction, Cell Walls, Non-Arabidopsis Systems, and Interactions with Biotic and Abiotic Factors. All conference abstracts and the full program are posted at The Arabidopsis Information Resource (TAIR), a publicly-accessibly website (www.arabidopsis.org/news/abstracts.jsp.) A survey completed by approximately 40% of the meeting attendees showed high satisfaction with the quality of the presentations, meeting organization and the city of Montreal. The conference is the largest annual international Arabidopsis venue which allowed the exchange of information at the forefront of Arabidopsis research and facilitated the establishment of new, and the strengthening of old, collaborations and networks. In addition, the conference provides the site for the annual meetings of the Multinational Arabidopsis Steering Committee (MASC) and the North American Arabidopsis Steering Committee (NAASC.) Graduate students, postdoctoral researchers, junior faculty, and underrepresented minorities made up half of the oral presentations thereby promoting the training of young scientists and facilitating important career development opportunities for speakers. Several poster sessions provided an opportunity for younger participants to freely meet with more established scientists. The NAASC continued its outreach efforts and again sponsored two special luncheons to encourage personal and professional development of young scientists and underrepresented minorities. The Emerging Scientists Luncheon featured 8 graduate students selected on the basis of scientific excellence of their submitted research abstracts. Also attending were the Keynote Speaker and faculty conference organizers. The Minority Funding Luncheon, featured 7 awardees (2 female graduate students, 2 female faculty, 2 male graduate students and 1 male faculty) selected by the NAASC through a widely-publicized application process. This luncheon was established specifically to provide an opportunity for underrepresented minorities, and/or scientists from Minority-Serving Institutions/Historically Black Colleges and Universities to network with NAASC members and representatives from federal funding agencies in an informal and intimate setting. This luncheon included introductions of each award recipient and discussion of outreach efforts and informal research and career discussions. Staff members from The Arabidopsis Information Resource (TAIR), the public U.S. Arabidopsis bioinformatics resource, led one workshop and participated in another to convey information to the community about Arabidopsis resources. Participation by young researchers was facilitated through DOE-sponsored registration awards to 10 early career applicants from the US including five graduate students, four postdoctoral scholars and one new assistant faculty member.

59 BASIC BIOLOGICAL SCIENCES↗

Investigation of psychophysical similarity measures for selection of similar images in the diagnosis of clustered microcalcifications on mammograms

The presentation of images with lesions of known pathology that are similar to an unknown lesion may be helpful to radiologists in the diagnosis of challenging cases for improving the diagnostic accuracy and also for reducing variation among different radiologists. The authors have been developing a computerized scheme for automatically selecting similar images with clustered microcalcifications on mammograms from a large database. For similar images to be useful, they must be similar from the point of view of the diagnosing radiologists. In order to select such images, subjective similarity ratings were obtained for a number of pairs of clustered microcalcifications by breast radiologists for establishment of a ''gold standard'' of image similarity, and the gold standard was employed for determination and evaluation of the selection of similar images. The images used in this study were obtained from the Digital Database for Screening Mammography developed by the University of South Florida. The subjective similarity ratings for 300 pairs of images with clustered microcalcifications were determined by ten breast radiologists. The authors determined a number of image features which represent the characteristics of clustered microcalcifications that radiologists would use in their diagnosis. For determination of objective similarity measures, an artificial neural network (ANN) was employed. The ANN was trained with the average subjective similarity ratings as teacher and selected image features as input data. The ANN was trained to learn the relationship between the image features and the radiologists' similarity ratings; therefore, once the training was completed, the ANN was able to determine the similarity, called a psychophysical similarity measure, which was expected to be close to radiologists' impressions, for an unknown pair of clustered microcalcifications. By use of a leave-one-out test method, the best combination of features was selected. The correlation coefficient between the gold standard and the psychophysical similarity measure through the use of seven features was relatively high (r=0.71) and was comparable to the correlation coefficients between the ratings by one radiologist and the average ratings by nine radiologists (r=0.69{+-}0.07). The correlation coefficient was improved compared to that of a distance-based method (r=0.58). The result indicated that similar images selected by the psychophysical similarity measure may be useful to radiologists in the diagnosis of clustered microcalcifications on mammograms.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Prospecting Biotechnologically-Relevant Monooxygenases from Cold Sediment Metagenomes: An In Silico Approach

The goal of this work was to identify sequences encoding monooxygenase biocatalysts with novel features by in silico mining an assembled metagenomic dataset of polar and subpolar marine sediments. The targeted enzyme sequences were Baeyer–Villiger and bacterial cytochrome P450 monooxygenases (CYP153). These enzymes have wide-ranging applications, from the synthesis of steroids, antibiotics, mycotoxins and pheromones to the synthesis of monomers for polymerization and anticancer precursors, due to their extraordinary enantio-, regio-, and chemo- selectivity that are valuable features for organic synthesis. Phylogenetic analyses were used to select the most divergent sequences affiliated to these enzyme families among the 264 putative monooxygenases recovered from the ~14 million protein-coding sequences in the assembled metagenome dataset. Three-dimensional structure modeling and docking analysis suggested features useful in biotechnological applications in five metagenomic sequences, such as wide substrate range, novel substrate specificity or regioselectivity. Further analysis revealed structural features associated with psychrophilic enzymes, such as broader substrate accessibility, larger catalytic pockets or low domain interactions, suggesting that they could be applied in biooxidations at room or low temperatures, saving costs inherent to energy consumption. This work allowed the identification of putative enzyme candidates with promising features from metagenomes, providing a suitable starting point for further developments.

59 BASIC BIOLOGICAL SCIENCES↗

Performance Comparison of Machine Learning Models for Ultrasonic Nondestructive Evaluation of Alkali-Silica Reaction in Concrete

Alkali-silica reaction (ASR) causes concrete degradation, leading to cracking, rebar corrosion, and reduced structural integrity, which raises safety concerns. Ultrasonic nondestructive evaluation (NDE) effectively assesses concrete properties and monitors ASR progression. However, its deployment and analysis require specialized expertise and subjective interpretation. As computational power increases, artificial intelligence (AI) and machine learning (ML) algorithms are increasingly being used to automate NDE data analysis across various industries for AI-assisted automation. Regulatory agencies are adapting to this technological shift, prompting a need to evaluate current ML technologies’ capabilities and limitations in assessing concrete material properties and damage. This report presents a comparative analysis of four ML regression models for predicting concrete material damage induced by ASR expansion using long-term ultrasonic data monitoring. The models investigated include linear regression (LR), support vector regression (SVR), shallow neural networks (NN), and deep neural networks (DNN). LR, SVR, and shallow NN models use features extracted from ultrasonic signals, whereas the DNN model processes time-domain ultrasonic signals and frequency spectra directly. The study systematically compared the models’ performance from various perspectives, including model input, prediction performance, and generalization ability. The findings indicate significant variability in model performance, with some ML algorithms achieving very high or very low prediction accuracy depending on the preprocessing and feature engineering (extraction and selection) applied. Key insights include the observation that shallow ML models (LR, SVR, and shallow NNs) require meticulous preprocessing and feature extraction to achieve high accuracy. In contrast, the DNN model, although it bypasses the need for feature engineering, necessitates extensive preprocessing to mitigate noise and computational demands. The SVR model emerged as the top performer among the shallow models, and the DNN model exhibited superior performance on specific datasets but struggled with generalization across specimens from different batches. Additionally, the SVR model is sensitive to temperature variations, whereas the DNN model is robust in this regard. Using recurrent neural networks is recommended for future ASR expansion prediction studies. Recurrent neural networks’ inherent ability to capture temporal dependencies and long-term patterns makes them well suited for analyzing sequential ultrasonic monitoring data. Overall, the results and conclusions of this study could provide insights into the capabilities and effectiveness of ML when applied to ultrasonic NDE data and help identify best practices for using ML for ultrasonic NDE of concrete material properties.

36 MATERIALS SCIENCE↗