Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Operator regression”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

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

At least 559 records · Page 31
All sourcesNASA NTRSDOE OSTI
DOE OSTI2025

Enhancing Automotive Intrusion Detection Through Multi-Modal Fusion: A CAN FD-LiDAR Approach

As vehicles become smarter and more autonomous, they increasingly depend on advanced sensors and communication technologies to operate securely. However, such growing dependence on technology—whether it’s CAN (Controller Area Network) for internal communication or LiDAR (Light Detection and Ranging) for sensing the world around them—also expands the attack surface for the types of cyber attacks. Traditional intrusion detection systems (IDS) typically monitor these systems in isolation, limiting their ability to detect sophisticated, crosssystem attacks. To address this, we propose a multi-modal fusion approach that combines real-world CAN FD signals (from the HCRL dataset) with LiDAR features (from the nuScenes dataset) to enhance attack detection. Our method employs a twostage ensemble approach. Calibrated XGBoost and LightGBM models initially process CAN FD (Fuzzing Data) and LiDAR data independently, detecting timing anomalies and space abnormalities. They are subsequently logarithmically combined with a logistic regression meta-model along with 17 engineered features capturing cross-modal behavior, prediction conflicts, and nonlinear interactions. This approach achieves an AUC of 0.87 and an F1-score of 0.82, surpassing single-modality baselines and early fusion methods, at merely 2 ms inference latency. Compared with deep learning competitors, it is 3 times more efficient, providing a lightweight, interpretable, and real time solution to automotive cybersecurity.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2020

Multi-objective optimization of peel and shear strengths in ultrasonic metal welding using machine learning-based response surface methodology

Ultrasonic metal welding (UMW) is a solid-state joining technique with varied industrial applications. Despite of its numerous advantages, UMW has a relative narrow operating window and is sensitive to variations in process conditions. As such, it is imperative to quantitatively characterize the influence of welding parameters on the resulting joint quality. The quantification model can be subsequently used to optimize the parameters. Conventional response surface methodology (RSM) usually employs linear or polynomial models, which may not be able to capture the intricate, nonlinear input-output relationships in UMW. Furthermore, some UMW applications call for simultaneous optimization of multiple quality indices such as peel strength, shear strength, electrical conductivity, and thermal conductivity. To address these challenges, this paper develops a machine learning (ML)-based RSM to model the input-output relationships in UMW and jointly optimize two quality indices, namely, peel and shear strengths. The performance of various ML methods including spline regression, Gaussian process regression (GPR), support vector regression (SVR), and conventional polynomial regression models with different orders is compared. A case study using experimental data shows that GPR with radial basis function (RBF) kernel and SVR with RBF kernel achieve the best prediction accuracy. The obtained response surface models are then used to optimize a compound joint strength indicator that is defined as the average of normalized shear and peel strengths. In addition, the case study reveals different patterns in the response surfaces of shear and peel strengths, which has not been systematically studied in the literature. While developed for the UMW application, the method can be extended to other manufacturing processes.

42 ENGINEERING↗
NASA NTRS2011

Sensitivity of C-Band Polarimetric Radar-Based Drop Size Distribution Measurements to Maximum Diameter Assumptions

The estimation of rain drop size distribution (DSD) parameters from polarimetric radar observations is accomplished by first establishing a relationship between differential reflectivity (Z(sub dr)) and the central tendency of the rain DSD such as the median volume diameter (D0). Since Z(sub dr) does not provide a direct measurement of DSD central tendency, the relationship is typically derived empirically from rain drop and radar scattering models (e.g., D0 = F[Z (sub dr)] ). Past studies have explored the general sensitivity of these models to temperature, radar wavelength, the drop shape vs. size relation, and DSD variability. Much progress has been made in recent years in measuring the drop shape and DSD variability using surface-based disdrometers, such as the 2D Video disdrometer (2DVD), and documenting their impact on polarimetric radar techniques. In addition to measuring drop shape, another advantage of the 2DVD over earlier impact type disdrometers is its ability to resolve drop diameters in excess of 5 mm. Despite this improvement, the sampling limitations of a disdrometer, including the 2DVD, make it very difficult to adequately measure the maximum drop diameter (D(sub max)) present in a typical radar resolution volume. As a result, D(sub max) must still be assumed in the drop and radar models from which D0 = F[Z(sub dr)] is derived. Since scattering resonance at C-band wavelengths begins to occur in drop diameters larger than about 5 mm, modeled C-band radar parameters, particularly Z(sub dr), can be sensitive to D(sub max) assumptions. In past C-band radar studies, a variety of D(sub max) assumptions have been made, including the actual disdrometer estimate of D(sub max) during a typical sampling period (e.g., 1-3 minutes), D(sub max) = C (where C is constant at values from 5 to 8 mm), and D(sub max) = M*D0 (where the constant multiple, M, is fixed at values ranging from 2.5 to 3.5). The overall objective of this NASA Global Precipitation Measurement Mission (GPM/PMM Science Team)-funded study is to document the sensitivity of DSD measurements, including estimates of D0, from C-band Z(sub dr) and reflectivity to this range of D(sub max) assumptions. For this study, GPM Ground Validation 2DVD's were operated under the scanning domain of the UAHuntsville ARMOR C-band dual-polarimetric radar. Approximately 7500 minutes of DSD data were collected and processed to create gamma size distribution parameters using a truncated method of moments approach. After creating the gamma parameter datasets the DSD's were then used as input to a T-matrix model for computation of polarimetric radar moments at C-band. All necessary model parameterizations, such as temperature, drop shape, and drop fall mode, were fixed at typically accepted values while the D(sub max) assumption was allowed to vary in sensitivity tests. By hypothesizing a DSD model with D(sub max) (fit) from which the empirical fit to D0 = F[Z(sub dr)] was derived via non-linear least squares regression and a separate reference DSD model with D(sub max) (truth), bias and standard error in D0 retrievals were estimated in the presence of Z(sub dr) measurement error and hypothesized mismatch in D(sub max) assumptions. Although the normalized standard error for D0 = F[Z(sub dr)r] can increase slightly (as much as from 11% to 16% for all 7500 DSDs) when the D(sub max) (fit) does not match D(sub max) (truth), the primary impact of uncertainty in D(sub max) is a potential increase in normalized bias error in D0 (from 0% to as much as 10% over all 7500 DSDs, depending on the extent of the mismatch between D(sub max) (fit) and D(sub max) (truth)). For DSDs characterized by large Z(sub dr) (Z(sub dr) > 1.5 to 2.0 dB), the normalized bias error for D0 estimation at C-band is sometimes unacceptably large (> 10%), again depending on the extent of the hypothesized D(sub max) mismatch. Modeled errors in D0 retrievals from Z(sub dr) at C-band are demonstrated in detail and comparedo similar modeled retrieval errors at S-band and X-band where the sensitivity to D(sub max) is expected to be less. The impact of D(sub max) assumptions to the retrieval of other DSD parameters such as Nw, the liquid water content normalized intercept parameter, are also explored. Likely implications for DSD retrievals using C-band polarimetric radar for GPM are assessed by considering current community knowledge regarding D(sub max) and quantifying the statistical distribution of Z(sub dr) from ARMOR over a large variety of meteorological conditions. Based on these results and the prevalence of C-band polarimetric radars worldwide, a call for more emphasis on constraining our observational estimate of D(sub max) within a typical radar resolution volume is made

Carey, Lawrence D.↗
NASA NTRS2020

Adaptive Stress Testing: Finding Likely Failure Events with Reinforcement Learning

Finding the most likely path to a set of failure states is important to the analysis of safety-critical systems that operate over a sequence of time steps, such as aircraft collision avoidance systems and autonomous cars. In many applications such as autonomous driving, failures cannot be completely eliminated due to the complex stochastic environment in which the system operates.As a result, safety validation is not only concerned about whether a failure can occur, but also discovering which failures are most likely to occur. This article presents adaptive stress testing (AST), a framework for finding the most likely path to a failure event in simulation. We consider a general black box setting for partially observable and continuous-valued systems operating in an environment with stochastic disturbances. We formulate the problem as a Markov decision process and use reinforcement learning to optimize it. The approach is simulation-based and does not require internal knowledge of the system, making it suitable for black-box testing of large systems. We present different formulations depending on whether the state is fully observable or partially observable. In the latter case, we present a modified Monte Carlo tree search algorithm that only requires access to the pseudorandom number generator of the simulator to overcome partial observability. We also present an extension of the framework, called differential adaptive stress testing (DAST), that can find failures that occur in one system but not in another. This type of differential analysis is useful in applications such as regression testing, where we are concerned with finding areas of relative weakness compared to a baseline. We demonstrate the effectiveness of the approach on an aircraft collision avoidance application, where a prototype aircraft collision avoidance system is stress tested to find the most likely scenarios of near mid-air collision.

Verification and Validation↗
DOE OSTI2020

Long Term Assessment of Radiocesium in Deer, Fish, Soil, and Vegetation at the Savannah River Site

Throughout the Savannah River Site's (SRS) operational history, many radionuclides have been released into the environment. However, due to fallout from weapons testing, the amount released from SRS, its persistence in the environment, and the seclusion of the Site, radiocesium (Cs-137) is one of the most critical radionuclides in the SRS environment. The Effective half-life (Te) is the time it takes for a radionuclide in the environment to decrease by 50% of its initial amount. There are three means of radionuclide removal: Physical: Radioactive decay, sedimentation, and washout; Chemical: Changes in pH, oxidation state or adsorption; Biological: Changes in the food web or translocation. Effective half-life (T{sub e}) = log{sub e}2/λ{sub e} Where λ{sub e} is estimated from the slopes of loge-transformed Cs-137 activity concentrations regressed on year. The objective of this work was to calculate the effective half-life of Cs-137 in various biota at SRS and to perform an assessment of the changes in the effective half life over time. Compared to 2013 data, there is a slight increase in the effective half-lives for fish at most sampling locations. This effective half-life increase is an indication that the rate of Cs-137 loss is starting to plateau. The effective half-life of Cs-137 in SRS fish ranges from 4-11 years. The effective half-life of Cs-137 in soil at the Creek plantation is about 15 years. The effective half-life of Cs-137 in grassy vegetation at the Creek Plantation is about 14 years. The effective half-life of deer hunted on the SRS is about 15.5 years. After termination of all reactor activity by 1980, the Cs-137 activity began to drop drastically.

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗
DOE OSTI2022

Impact of acetic acid exposure on metal contact degradation of different crystalline silicon solar cell technologies

Degradation due to acetic acid in photovoltaic (PV) modules has been a commonly observed phenomenon for both damp-heat exposure and outdoor operations. Acetic acid is a degradation byproduct of ethylene-vinyl acetate (EVA), a common module encapsulant. To address this issue, robust metallization pastes and cell technologies are being developed. However, it is important to assess how these technologies perform in an acetic acid environment and withstand degradation before they are implemented in the solar market. In this work, we investigate the impact of acetic acid exposure on four different cell groups: monofacial passivated emitter and rear contact (PERC) cells with advanced telluride-based front contact pastes, bifacial PERC cells with novel aluminum rear contact pastes, bifacial tunnel oxide passivated contacts (TOPCon) cells, and silicon heterojunction (SHJ) cells. These cells were exposed to acetic acid for different time increments. The recombination losses were characterized by Suns-VOC, and multi-variate regression analysis of intensity-dependent photoluminescence (PL) images with Griddler AI. Resistive losses were tracked with the transmission line method (TLM). Samples showing severe performance degradation were selected for further materials characterization to understand the root cause. Top-down and cross-sectional scanning electron microscopy (SEM), energy dispersive spectroscopy (EDS), and X-ray photoelectron spectroscopy (XPS) were performed to investigate the change in materials properties. Our study shows that the front contacts of the bifacial TOPCon cells and monofacial PERC cells were significantly affected by acetic acid exposure. Here, the SHJ cells were found to be the most stable.

14 SOLAR ENERGY↗
NASA NTRS1977

The effects of participatory mode and task workload on the detection of dynamic system failures

The ability of operators to detect step changes in the dynamics of control systems is investigated as a joint function of, (1) participatory mode: whether subjects are actively controlling those dynamics or are monitoring an autopilot controlling them, and (2) concurrent task workload. A theoretical analysis of detection in the two modes identifies factors that will favor detection in either mode. Three subjects detected system failures in either an autopilot or manual controlling mode, under single-task conditions and concurrently with a subcritical tracking task. Latency and accuracy of detection were assessed and related through a speed accuracy tradeoff representation. It was concluded that failure detection performance was better during manual control than during autopilot control, and that the extent of this superiority was enhanced as dual-task load increased. Ensemble averaging and multiple regression techniques were then employed to investigate the cues utilized by the subjects in making their detection decisions.

Wickens, C. D.↗
DOE OSTI2024

Development and Refinement of Bison Models for Higher Burnup Application

Electric Power Research Institute (EPRI) is an international research organization funded by organizations in the energy and government sectors. The company conducts research to tackle the energy challenges of its member organizations. In support of the U.S. nuclear industry, EPRI and Idaho National Laboratory were awarded a U.S. Department of Energy Technology Commercialization Fund to develop a loss of coolant accident (LOCA) analysis tool. This tool was developed to support burnup extension activities by supplementing the existing capabilities of operators of existing large light-water reactor plants. The tool couples INL developed codes Bison and RELAP5-3D for advanced simulation of LOCA behavior. The project consists of seven major tasks, one of which is the development and refinement of Bison models for higher burnup applications. This task is broken down into six targeted areas of development and refinement including: formation of high burnup structure, fuel fracture and pulverization, transient fission gas release, axial gas communication within a fuel rod, cladding ballooning and burst, and axial fuel relocation. This report summarizes the models for both the fuel and cladding in Bison for high burnup application, identifies the areas of development and refinement necessary to enhance the Bison capabilities to be used in the coupled simulation tool with RELAP5-3D. Integration and regression testing to demonstrate proper implementation of the new features in Bison is highlighted.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗
DOE OSTI2025

Sizing Accuracy of Low-Cost Optical Particle Sensors Under Controlled Laboratory Conditions

Low-cost particulate matter sensors have seen increased use for monitoring at personal and local levels due to their affordability, ease of operation, and high time resolution. However, the quality of data reported by these sensors can be questionable, and a thorough evaluation of their performance is necessary. This study evaluated the particle sizing accuracy of several commonly used optical sensors, including the Alphasense optical particle counter (OPC), TSI DustTrak DRX aerosol monitor, Plantower PMS5003 sensor, and Sensirion SPS30 sensor, using laboratory-generated monodisperse particles. The OPC and DRX agreed partially with reference instruments and showed promise in detecting coarse-size particles. However, the PMS5003 and SPS30 did not correctly size fine and coarse particles. Furthermore, their reported mass distributions do not directly correspond to their number distribution. Despite these limitations, field measurements involving a dust storm period showed that the SPS30 correlated reasonably well with reference instruments for both PM2.5 and PM10, though the regression slopes differed significantly. These findings underscore the need for caution when interpreting data from low-cost optical sensors, particularly for coarse particles. Recommendations for improving the performance of these sensors are also provided.

Gautam, Prakash↗
DOE OSTI2020

An Exploratory Approach Using Regression and Machine Learning in the Analysis of Mass Absorption Cross Section of Black Carbon Aerosols: Model Development and Evaluation

Mass absorption cross-section of black carbon (MAC BC ) describes the absorptive cross-section per unit mass of black carbon, and is, thus, an essential parameter to estimate the radiative forcing of black carbon. Many studies have sought to estimate MAC BC from a theoretical perspective, but these studies require the knowledge of a set of aerosol properties, which are difficult and/or labor-intensive to measure. We therefore investigate the ability of seven data analytical approaches (including different multivariate regressions, support vector machine, and neural networks) in predicting MAC BC for both ambient and biomass burning measurements. Our model utilizes multi-wavelength light absorption and scattering as well as the aerosol size distributions as input variables to predict MAC BC across different wavelengths. We assessed the applicability of the proposed approaches in estimating MAC BC using different statistical metrics (such as coefficient of determination (R 2 ), mean square error (MSE), fractional error, and fractional bias). Overall, the approaches used in this study can estimate MAC BC appropriately, but the prediction performance varies across approaches and atmospheric environments. Based on an uncertainty evaluation of our models and the empirical and theoretical approaches to predict MAC BC , we preliminarily put forth support vector machine (SVM) as a recommended data analytical technique for use. We provide an operational tool built with the approaches presented in this paper to facilitate this procedure for future users.

54 ENVIRONMENTAL SCIENCES↗
NASA NTRS2023

Interpretation of Probabilistic Surface Ozone Forecasts: A Case Study for Philadelphia

The use of probabilistic forecasting has been growing in a variety of disciplines because of its potential to emphasize the degree of uncertainty inherent in a prediction. Interpretation of probabilistic forecasts, however, is oftentimes difficult, deterring users who may benefit from such forecasts. To encourage broader use of probabilistic forecasts in the field of air quality, a process for interpreting forecasts from a statistical probabilistic air quality surface ozone model [the Regression in Self Organizing Map (REGiS)] is demonstrated. Four procedures to convert probabilistic to deterministic forecasts are explored for the Philadelphia, Pennsylvania, metropolitan area. These procedures calibrate the predicted probability of daily maximum 8-h-average ozone exceeding a standard value by 1) estimating climatological relative frequency, 2) establishing a probability of an exceedance threshold as 50%, 3) maximizing the threat score, and 4) determining the unit bias ratio. REGiS is trained using 2000–11 ozone-season (1 May–30 September) data, calibrated using 2012–14 data, and evaluated using 2015–18 data. Assessment of the calibration data with the Pierce skill score suggests an exceedance threshold based on climatological relative frequency for the conversion from probabilistic to deterministic forecasts. Calibrated REGiS generally compares well to predictions from the U.S. national air quality model and operational “expert” forecasts over the evaluation period. For other probabilistic models and situations, different procedures of converting probabilistic to deterministic forecasts may be more beneficial. The methods presented in this paper represent an approach for operational air quality forecasters seeking to use probabilistic model output to support forecasts designed to protect public health.

Nikolay Balashov↗
DOE OSTI2020

Nested active learning for efficient model contextualization and parameterization: pathway to generating simulated populations using multi-scale computational models

There is increasing interest in the use of mechanism-based multi-scale computational models (such as agent-based models (ABMs)) to generate simulated clinical populations in order to discover and evaluate potential diagnostic and therapeutic modalities. The description of the environment in which a biomedical simulation operates (model context) and parameterization of internal model rules (model content) requires the optimization of a large number of free parameters. In this work, we utilize a nested active learning (AL) workflow to efficiently parameterize and contextualize an ABM of systemic inflammation used to examine sepsis. Contextual parameter space was examined using four parameters external to the model’s rule set. The model’s internal parameterization, which represents gene expression and associated cellular behaviors, was explored through the augmentation or inhibition of signaling pathways for 12 signaling mediators associated with inflammation and wound healing. We have implemented a nested AL approach in which the clinically relevant (CR) model environment space for a given internal model parameterization is mapped using a small Artificial Neural Network (ANN). The outer AL level workflow is a larger ANN that uses AL to efficiently regress the volume and centroid location of the CR space given by a single internal parameterization. We have reduced the number of simulations required to efficiently map the CR parameter space of this model by approximately 99%. In addition, we have shown that more complex models with a larger number of variables may expect further improvements in efficiency.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS1980

Effects of primary rotor parameters on flapping dynamics

The effects of flapping dynamics of four main rotor design features that influence the agility, stability, and operational safety of helicopters are studied. The parameters include flapping hinge offset, flapping hinge restraint, pitch-flap coupling, and blade lock number. First, the flapping equations of motion are derived that explicitly contain the design parameters. The dynamic equations are then developed for the tip-path plane, and the influence of individual and combined variations in the design parameters determined. The steady state flapping response is examined with respect to control input and aircraft angular rate which leads to a feedforward control law for control decoupling through cross feed, and a feedback control law to decouple the steady state flapping response. The condition for achieving perfect decoupling of the flapping response due to aircraft pitch and roll rates without using feedback control is also found for the hover case. It is indicated that the frequency of the regressing flapping mode of the rotor system can become low enough to require consideration in the assessment of handling characteristics.

Chen, R. T. N.↗
DOE OSTI2020

Bayesian Approach to Estimation of Water Table Elevations Using Historical Rasters as Prior Information 2019 - 20430

In cases of complex but only partially known geology and a lack of spatial control in observation well locations, water table elevation estimation is very challenging. In some cases, auxiliary information, such as observations of the movement of tracers, operation of injection and extraction wells, and calibration of groundwater models against the historical elevation data, can be combined with expert judgement to estimate flow directions in areas of sparse data and to aid in the production of more reliable contour maps (and associated rasters) than could be produced by relying on sparse well elevation data alone. Given a historical sequence of these raster maps, the question arises how to automate, to the extent possible, the process of producing new raster maps to reflect data from previous times, the current data and the operation of expert judgement. One solution is to adopt a Bayesian point of view and to regard the historical well elevation data, auxiliary information and historical raster maps as prior information. The well elevations for water table wells, as well as those for injection/extraction wells and the data associated with other relevant variables, can be viewed as predictors for the raster surface. From this prior information, we can, conditional on the values of these predictors for a new time period, compute an expected value map and a standard deviation map for the new raster. These then can be taken to specify a prior predictive distribution for the pixels in the new raster map. Then we condition the pixels, corresponding to water level observation wells within the raster, on the observed values in those wells (which in general will differ from the regression estimate) for the new time period. Given the smoothness of the water table surface, we then smooth the surface of deviations from the mean surface, based on the variograms of the historical rasters, and add this smoothed surface to the regression mean surface. The error structure of the produced raster map is defined by the regression error structure and the error due to smoothing based on the estimated variograms. This methodology has been developed and is being further refined for groundwater monitoring and remediation at LANL. It is a very flexible method that can also be applied with a variety of other predictors applied to model the water level wells in the area of interest over the historical record. The smoothness of the spatial process and its possible evolution over time can then be estimated from the residuals from this regression. This can be augmented by expert hydrogeological opinion based on site topography and hydrogeology. (authors)

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2023

Evaluating the impact of wildfire smoke on solar photovoltaic production

There are growing needs to understand how extreme weather events impact the electrical grid. Renewable energy sources such as solar photovoltaics are expanding in use to help sustainably meet electricity demands. Wildfires and, notably, the widespread smoke resulting from them, are one such extreme event that can impair the performance of solar photovoltaics. However, isolating the impact that smoke has on photovoltaic energy production, separate from ambient conditions, can be difficult. In this work, we seek to understand and quantify the impacts of wildfire smoke on solar photovoltaic production within the Western United States. Our analysis focuses on the construction of a random forest regression model to predict overall solar photovoltaic production. The model is used to separate and quantify the impacts of wildfire smoke in particular. To do so, we fuse historical weather, solar photovoltaic energy production, and PM2.5 particulate matter (primary smoke pollutant) data to train and test our model. The additional weather data allows us to capture interactions between wildfire smoke and other ambient conditions, as well as to create a more powerful predictive model capable of better quantifying the impacts of wildfire smoke on its own. We find that solar PV energy production decreases 8.3% on average during high smoke days at PV sites as compared to similar conditions without smoke present. Finally, this work allows us to improve our understanding of the potential impact on photovoltaic-based energy production estimates due to wildfire events and can help inform grid and operational planning as solar photovoltaic penetration levels continue to grow.

14 SOLAR ENERGY↗
NASA NTRS2018

Trusted Communication: Utilizing Speech Communication to Enhance Human-Machine Teaming Success

An area of increasing interest for the next generation of aircraft is autonomy and the integration of increasingly autonomous systems into the national airspace. Such an integration requires humans to work closely with autonomous systems, forming teams. Our hypothesis is that a team composed of both humans and autonomous systems will operate better than either entity alone. We have existing procedures for certifying pilots to operate in the national airspace and are currently working on methods for validating the function of autonomous systems, however we have no method in place for assessing the interaction of these two disparate systems. Communication is one avenue. This paper will examine the use of language as a metric for ascertaining human-machine teaming effectiveness. A proof-of-concept of the application of two communication-based analysis techniques, Linguistic Inquiry and Word Count (LIWC) and Latent Semantic Analysis (LSA), for the prediction of success in human/chatbot teaming was conducted. By running these analyses over data from the 2014 and 2015 Loebner Prize competitions of human/chatbot teaming, numerical scores were obtained that can be associated with scores provided by human judges during the competition. Correlating their LIWC and LSA data with the scores provided by the judges, and using linear regression over this correlation, formulae were obtained that predict the score of human/chatbot interaction. These formulae were tested over the 2013 Loebner Prize transcripts, determining that, though there was strong correlation between predicted and actual scores, the predictive success of this method was not strong. However, with specialized topic spaces and lexica, as well as larger data sets, the predictive power of these metrics will improve. Given the importance of providing metrics for human-machine system team success and given the promise shown by the communication-basedLIWCand LSAmethods, continuing research in this area is necessary. After examining the potential for using communication and spoken language as a metric for the success of human/autonomous system teaming, this paper then examines aspects inherent to communication systems that may contribute to unreliability and reduced trust. Modern natural language processing tools rely on deep learning algorithms to create language rules that produce accurate results, but these rules are uninterpretable. The resulting blackbox system lacks transparency necessary for full validation and complete trust. Additionally, speech-based interfaces pose other difficulties to developing coordinated teamwork between humans and autonomous systems. Human communication is infrequently limited to speech only, instead usually relying on a combination of verbal, gestural, and general body language communication. Reducing an analysis of team effectiveness to a study of spoken language alone is problematic as it leaves these other equally important forms of communication out. This paper will examine these problems and the general deficiencies in speech-based metrics for human-machine teaming.

E L Meszaros↗
DOE OSTI2023

Residuals-based distributionally robust optimization with covariate information

We consider data-driven approaches that integrate a machine learning prediction model within distributionally robust optimization (DRO) given limited joint observations of uncertain parameters and covariates. Our framework is flexible in the sense that it can accommodate a variety of regression setups and DRO ambiguity sets. We investigate asymptotic and finite sample properties of solutions obtained using Wasserstein, sample robust optimization, and phi-divergence-based ambiguity sets within our DRO formulations, and explore cross-validation approaches for sizing these ambiguity sets. Through numerical experiments, we validate our theoretical results, study the effectiveness of our approaches for sizing ambiguity sets, and illustrate the benefits of our DRO formulations in the limited data regime even when the prediction model is misspecified.

97 MATHEMATICS AND COMPUTING↗