Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Inference”

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 487 records · Page 27

Risk-Informed Condition Evaluation of Solar-centered Energy Generation and Distribution Networks through Bayesian Learning and Inference

We develop a methodology based on Bayesian inference over Probabilistic Graphical Models (PGMs) to understand and quantify risk in solar-centered grids using targeted measurements and learned system behavior. Being non-prescriptive but, rather, able to infer system behavior and, ultimately, address risk queries from data, our machine learning-type paradigm is tailored for diverse topologies and threat scenarios often associated with distributed energy generation and photovoltaic distributed energy resources (PV-DERs) in particular. We describe algorithmic processes for: (i) learning the structure of PGMs that result from attack-prone PV-DER-proliferated distribution systems, (ii) quantifying cause-effect relationships, and (iii) evaluating risk queries based on diverse evidence. The contributions are illustrated on a residential grid subject to output impairment attacks on its PV-DER infrastructure.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Estimating Travel Energy Consumption Uncertainty Based on Inferred Travel Mode and Sensed Travel Length

To properly inform transport policy and infrastructure changes, transportation related metrics need both measured values and uncertainties of those values. Travel monitoring smartphone apps can record people's travel behavior, but trip data quality is limited by sensor errors, user labeling rates and the accuracy of inference algorithms used for travel diary creation. We discuss the use of phone app recorded travel diary data to estimate energy consumption, and propose the use of propagation of variance to find error bars for such estimates. We define energy consumption for one trip as trip length times the energy intensity per distance unit of the travel mode used. We characterize trip length errors with relative error and inferred trip mode errors with confusion matrix columns. The resulting variances of each measurement are then propagated to the final calculated energy consumption. We tested our uncertainty methods on a dataset that used phone app data combined with prompted recall, consisting of 92,234 labeled trips for over 500,000 miles. Accounting for uncertainty using expected energy intensities and variance propagation gives a dataset-wide aggregate energy consumption percent error of about 8%, within one standard deviation from the truth. Future work could involve applying similar methods to other travel diary based metrics.

ADVANCED PROPULSION SYSTEMS,ENERGY PLANNING, POLIC↗

Constraining Cosmology with Simulation-based inference and Optical Galaxy Cluster Abundance

We test the robustness of simulation-based inference (SBI) in the context of cosmological parameter estimation from galaxy cluster counts and masses in simulated optical datasets. We construct ``simulations'' using analytical models for the galaxy cluster halo mass function (HMF) and for the observed richness (number of observed member galaxies) to train and test the SBI method. We compare the SBI parameter posterior samples to those from an MCMC analysis that uses the same analytical models to construct predictions of the observed data vector. The two methods exhibit comparable performance, with reliable constraints derived for the primary cosmological parameters, ($\Omega_m$ and $\sigma_8$), and richness-mass relation parameters. We also perform out-of-domain tests with observables constructed from galaxy cluster-sized halos in the Quijote simulations. Again, the SBI and MCMC results have comparable posteriors, with similar uncertainties and biases. Unsurprisingly, upon evaluating the SBI method on thousands of simulated data vectors that span the parameter space, SBI exhibits worsened posterior calibration metrics in the out-of-domain application. We note that such calibration tests with MCMC is less computationally feasible and highlight the potential use of SBI to stress-test limitations of analytical models, such as in the use for constructing models for inference with MCMC.

79 ASTRONOMY AND ASTROPHYSICS↗

Assessing Membership Inference Attacks under Distribution Shifts

Membership inference attacks (MIAs) exploit machine learning models to infer whether a data point was in the training set, posing significant privacy risks even with limited black-box access. These attacks rely on the attacker approximating the target model’s training distribution, yet the impact of distribution shifts between target and shadow models on MIA success remains underexplored. We systematically evaluate five types of distribution shifts —-cutout, jitter, Gaussian noise, label shift, and attribute shift —- at varying intensities. Our results reveal that these shifts affect MIA effectiveness in nuanced ways, with some reducing attack success while others exacerbate vulnerabilities, and the same shift can have opposite effects depending on the type of MIA. This highlights the complex interplay between distributional differences and attack performance, offering critical insights for improving model defenses against MIAs.

Shi, Yichuan [Massachusetts Institute of Technolog↗

Simulation-based inference for neutrino interaction model parameter tuning

High-energy physics experiments studying neutrinos rely heavily on simulations of their interactions with atomic nuclei. Limitations in the theoretical understanding of these interactions typically necessitate ad hoc tuning of simulation model parameters to data. Traditional tuning methods for neutrino experiments have largely relied on simple algorithms for numerical optimization. While adequate for the modest goals of initial efforts, the complexity of future neutrino tuning campaigns is expected to increase substantially, and new approaches will be needed to make progress. In this paper, we examine the application of simulation-based inference (SBI) to the neutrino interaction model tuning for the first time. Using a previous tuning study performed by the MicroBooNE experiment as a test case, we find that our SBI algorithm can correctly infer the tuned parameter values when confronted with a mock data set generated according to the MicroBooNE procedure. This initial proof-of-principle illustrates a promising new technique for next-generation simulation tuning campaigns for the neutrino experimental community.

Tame-Narvaez, Karla [Fermilab] (ORCID:000000022249↗

MoE-Inference-Bench: Performance Evaluation of Mixture of Expert Large Language and Vision Models

Mixture of Experts (MoE) models have enabled the scaling of Large Language Models (LLMs) and Vision Language Models (VLMs) by achieving massive parameter counts while maintaining computational efficiency. However, MoEs introduce several inference-time challenges, including load imbalance across experts and the additional routing computational overhead. To address these challenges and fully harness the benefits of MoE, a systematic evaluation of hardware acceleration techniques is essential. We present MoE-Inference-Bench, a comprehensive study to evaluate MoE performance across diverse scenarios. We analyze the impact of batch size, sequence length, and critical MoE hyperparameters such as FFN dimensions and number of experts on throughput. We evaluate several optimization techniques on Nvidia H100 GPUs, including pruning, Fused MoE operations, speculative decoding, quantization, and various parallelization strategies. Our evaluation includes MoEs from the Mixtral, DeepSeek, OLMoE and Qwen families. The results reveal performance differences across configurations and provide insights for the efficient deployment of MoEs.

Chitty-Venkata, Krishna Teja↗

Approach for Inferring Full-Scope Human Reliability Data Based on Simplified Simulator Data

This paper proposes a method for inferring full-scope human reliability data based on the Simplified Human Error Experimental Program (SHEEP) data. It mainly focuses on the human errors observed when using simulators with different complexity levels. In the proposed method, the manner in which human error probabilities (HEPs) change as a result of increasing simulator complexity and how simulator complexity levels are quantified represent key information for inferring full-scope data. In the present study, SHEEP error data pertaining to actual professional operators using Rancor Microworld (Rancor) (i.e., a more simplified simulator) and Compact Nuclear Simulator (CNS) (i.e., a less simplified simulator) were compared with the HuREX error data. An approach to quantifying simulator complexity levels was then proposed based on information theory and acquired eye-tracker data.

99 - GENERAL AND MISCELLANEOUS↗

Forecasting Dark Matter Subhalo Constraints from Stellar Streams using Implicit Likelihood Inference

The evidence for dark matter (DM) remains compelling, although attempts to understand its particle nature remain inconclusive. One promising method to study DM is detecting DM subhalos through their gravitational interactions with stellar streams. In this study, we apply Neural Posterior Estimation (NPE) to constrain subhalo interaction parameters, including mass, scale radius, velocity, and encounter geometry, from stellar stream kinematics. We generate particle spray simulations based on the Lagrange Cloud stripping technique, focusing on the ATLAS-Aliqa Uma stream as a test case. We train multiple NPE models across multiple observational scenarios, quantifying how kinematic completeness affects inference and forecasting constraints from upcoming surveys including LSST, 4MOST, and 10-year Gaia data. Our results demonstrate that NPE can produce accurate and well-calibrated posteriors. In the idealized case with full 6D coordinates, we achieve subhalo mass uncertainties of 15-20% for a $10^7 \, \mathrm{M_\odot}$ subhalo, with 5D coordinates (excluding radial velocities) achieving similar performance. Under realistic observational conditions, mass uncertainties range from 50% (present-day) to 20-40% (future scenarios), with comparable performance between the photometric-only LSST sample and a smaller sample that includes Gaia proper motions and 4MOST radial velocities. Most notably, we find that velocity bimodality emerges when phase space is poorly sampled, whether due to missing kinematic information or limited stellar tracers. Combining large photometric samples with targeted spectroscopic follow-up can effectively resolves this degeneracy. These results demonstrate the power of implicit likelihood inference for optimizing stellar stream observational strategies and forecasting DM subhalo constraints from upcoming surveys.

Nguyen, Tri [Northwestern U. (main); SkAI, Chicago↗

First Estimation of Model Parameters for Neutrino-Induced Nucleon Knockout Using Simulation-Based Inference

To enable an accurate determination of oscillation parameters, accelerator-based neutrino experiments require detailed simulations of nuclear interaction physics in the GeV regime. While substantial effort from both theory and experiment is currently being invested to improve the fidelity of these simulations, their present deficiencies typically oblige experimental collaborations to resort to empirical tuning of simulation model parameters. As the precision requirements of the field continue to become more stringent, machine learning techniques may provide a powerful means of handling corresponding growth in the complexity of future neutrino interaction model tuning exercises. To study the suitability of simulation-based inference (SBI) for this physics application, in this paper we revisit a tuned configuration of the GENIE neutrino event generator that was originally developed by the MicroBooNE collaboration. Despite closely reproducing the adopted values of four physics parameters when confronted with the tuned cross-section predictions as input, we find that our trained SBI algorithm prefers modestly different values (within MicroBooNE's assigned uncertainties) and achieves slightly better goodness-of-fit when inference is run on the experimental data set originally used by MicroBooNE. We also find that our trained algorithm can create a fair approximation of an alternative neutrino scattering simulation, NuWro, that shares only a subset of its physics model parameters with GENIE.

Tame-Narvaez, Karla [Fermilab] (ORCID:000000022249↗

Exploration of Real Time Inference for MI-RR Deblending on GPU/TPU Systems

The Fermilab Main Injector (MI) and Recycler Ring (RR) share a common beam loss monitor (BLM) system, making loss events difficult to attribute to their source machine when beam is present in both simultaneously. The Real-time Edge AI for Distributed Systems (READS) project addresses this by deblending BLM readings in real time using machine learning (ML). The current FPGA based implementation meets the sub-3 ms latency requirement but carries a resource intensive hls4ml development cycle, motivating exploration of GPU based deployment. This paper characterizes inference latency on an NVIDIA Jetson Orin Nano and introduces a packet organization scheme for assembling synchronized event frames from seven distributed BLM DAQ streams. Using a Python based DAQ simulation with injected timing jitter in place of unavailable live beam data, the pipeline achieved an average end to end latency of 0.456 ms (σ = 0.122 ms) across 167,000 test frames, comfortably meeting the timing constraint. Early outliers were attributed to TensorRT warm-up rather than steady state limitations, suggesting GPU based inference is a viable alternative to the existing FPGA implementation.

Yu, Kellen [Fermilab; Cornell U.]↗

A new method for multinomial inference using Dempster-Shafer theory

A new method for multinomial inference is proposed by representing the cell probabilities as unordered segments on the unit interval and following Dempster-Shafer (DS) theory. The resulting DS posterior is then strengthened to improve symmetry and learning properties with the final posterior model being characterized by a Dirichlet distribution. In addition to computational simplicity, the new model has desirable invariance properties related to category permutations, refinements, and coarsenings. Furthemore, posterior inference on relative probabilities amongst certain cells depends only on data for the cells in question. Finally, the model is quite flexible with regard to parameterization and the range of testable assertions. Comparisons are made to existing methods and illustrated with two examples.

97 MATHEMATICS AND COMPUTING↗

Approximate inference on planar graphs using loop calculus and belief progagation

We introduce novel results for approximate inference on planar graphical models using the loop calculus framework. The loop calculus (Chertkov and Chernyak, 2006b) allows to express the exact partition function Z of a graphical model as a finite sum of terms that can be evaluated once the belief propagation (BP) solution is known. In general, full summation over all correction terms is intractable. We develop an algorithm for the approach presented in Chertkov et al. (2008) which represents an efficient truncation scheme on planar graphs and a new representation of the series in terms of Pfaffians of matrices. We analyze in detail both the loop series and the Pfaffian series for models with binary variables and pairwise interactions, and show that the first term of the Pfaffian series can provide very accurate approximations. The algorithm outperforms previous truncation schemes of the loop series and is competitive with other state-of-the-art methods for approximate inference.

97 MATHEMATICS AND COMPUTING↗

Nuclear Forensic Inferences Using Iterative Multidimensional Statistics

Nuclear forensics involves the analysis of interdicted nuclear material for specific material characteristics (referred to as 'signatures') that imply specific geographical locations, production processes, culprit intentions, etc. Predictive signatures rely on expert knowledge of physics, chemistry, and engineering to develop inferences from these material characteristics. Comparative signatures, on the other hand, rely on comparison of the material characteristics of the interdicted sample (the 'questioned sample' in FBI parlance) with those of a set of known samples. In the ideal case, the set of known samples would be a comprehensive nuclear forensics database, a database which does not currently exist. In fact, our ability to analyze interdicted samples and produce an extensive list of precise materials characteristics far exceeds our ability to interpret the results. Therefore, as we seek to develop the extensive databases necessary for nuclear forensics, we must also develop the methods necessary to produce the necessary inferences from comparison of our analytical results with these large, multidimensional sets of data. In the work reported here, we used a large, multidimensional dataset of results from quality control analyses of uranium ore concentrate (UOC, sometimes called 'yellowcake'). We have found that traditional multidimensional techniques, such as principal components analysis (PCA), are especially useful for understanding such datasets and drawing relevant conclusions. In particular, we have developed an iterative partial least squares-discriminant analysis (PLS-DA) procedure that has proven especially adept at identifying the production location of unknown UOC samples. By removing classes which fell far outside the initial decision boundary, and then rebuilding the PLS-DA model, we have consistently produced better and more definitive attributions than with a single pass classification approach. Performance of the iterative PLS-DA method compared favorably to that of classification and regression tree (CART) and k nearest neighbor (KNN) algorithms, with the best combination of accuracy and robustness, as tested by classifying samples measured independently in our laboratories against the vendor QC based reference set.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗

Bremsstrahlung and K(alpha) fluorescence measurements for inferring conversion efficiencies into fast ignition relevant hot electrons

The Bremsstrahlung and K-shell emission from 1 mm x 1 mm x 1 mm planar targets irradiated by a short-pulse 3 x 10{sup 18}-8 x 10{sup 19} W/cm{sup 2} laser were measured. The Bremsstrahlung was measured using a filter stack spectrometer with spectral discrimination up to 500 keV. K-shell emission was measured using a single photon counting charge coupled device (CCD). From Monte Carlo modeling of the target emission, conversion efficiencies into 1-3 MeV electrons of 3-12%, representing 20-40% total conversion efficiencies were inferred for intensities up to 8 x 10{sup 19} W/cm{sup 2}. Comparisons to scaling laws using synthetic energy spectra generated from the intensity distribution of the focal spot imply slope temperatures less than the ponderomotive potential of the laser. Resistive transport effects may result in potentials of a few hundred kV in the first few tens of microns in the target. This would lead to higher total conversion efficiencies than inferred from Monte Carlo modeling but lower conversion efficiencies into 1-3 MeV electrons.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Measurement of off-shell Higgs boson production in the $H^*\rightarrow ZZ\rightarrow 4\ell$ decay channel using a neural simulation-based inference technique in 13 TeV pp collisions with the ATLAS detector

A measurement of off-shell Higgs boson production in the $H^*\to ZZ\to 4\ell$ decay channel is presented. The measurement uses 140 fb −1 of proton–proton collisions at $\sqrt{s} = 13$ TeV collected by the ATLAS detector at the Large Hadron Collider and supersedes the previous result in this decay channel using the same dataset. The data analysis is performed using a neural simulation-based inference method, which builds per-event likelihood ratios using neural networks. The observed (expected) off-shell Higgs boson production signal strength in the $ZZ\to 4\ell$ decay channel at 68% CL is $0.87^{+0.75}_{-0.54}$ ($1.00^{+1.04}_{-0.95}$ ). The evidence for off-shell Higgs boson production using the $ZZ\to 4\ell$ decay channel has an observed (expected) significance of 2.5σ (1.3σ). The expected result represents a significant improvement relative to that of the previous analysis of the same dataset, which obtained an expected significance of 0.5σ. When combined with the most recent ATLAS measurement in the $ZZ\to 2\ell 2\nu$ decay channel, the evidence for off-shell Higgs boson production has an observed (expected) significance of 3.7σ (2.4σ). The off-shell measurements are combined with the measurement of on-shell Higgs boson production to obtain constraints on the Higgs boson total width. The observed (expected) value of the Higgs boson width at 68% CL is $4.3^{+2.7}_{-1.9}$ ($4.1^{+3.5}_{-3.4}$ ) MeV.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

A GPU‐Accelerated Generative Adversarial Model for Causal Inference

We develop a GPU-accelerated machine learning generative adversarial model designed to facilitate causal inferences from observational data. Our model's theoretical framework is conceptualized in a manner that is amenable to being operable and scalable for high-performance computing platforms. We leverage GPU acceleration to develop a parallel evolutionary algorithm to achieve large-scale parallel computation of the model within a now widely accessible computing platform. This capability both enhances computational speedup and efficiency and also extends the use of the model to a broader range of substantive research domains while maintaining the underlying theoretical properties of the model.

GPU↗

Capabilities of multivariate Bayesian inference toward seismic hazard assessment

Multivariate Bayesian analysis can bring significant benefits to seismic hazard analysis: Its multivariate feature enables computing scalar and vector hazard without making any approximations; Correlations between intensity measures are implicitly modeled, permitting direct simulation of ground motion selection tools such as the conditional mean spectrum and the generalized conditioning intensity measure; and Its updating feature enables a seamless integration of new ground motion data into the hazard results. Here, we first develop a multivariate Bayesian ground motion model through the NGA-West2 database. The model functional form considers fault-type, magnitude, and distance dependencies, and also the linear and the rock intensity dependent site response. We use a hybrid Markov Chain Monte Carlo sampling to perform Bayesian inference consisting of Gibbs step and a multilevel Metropolis-Hastings step. We then perform several checks on the model and note that its performance is satisfactory. Finally, we illustrate the merits of this multivariate Bayesian analysis, which include: ground motion model updating with ground motion data recorded in the last four years not part of the NGA-West2 database; computation of scalar and vector seismic hazard using the un-updated and updated ground motion models for Los Angeles, CA; and simulation of the conditional mean spectrum under scalar and vector IM conditioning while accounting for different sources of aleatoric and epistemic uncertainties.

58 GEOSCIENCES↗

Data free inference with processed data products

Here, we consider the context of probabilistic inference of model parameters given error bars or confidence intervals on model output values, when the data is unavailable. We introduce a class of algorithms in a Bayesian framework, relying on maximum entropy arguments and approximate Bayesian computation methods, to generate consistent data with the given summary statistics. Once we obtain consistent data sets, we pool the respective posteriors, to arrive at a single, averaged density on the parameters. This approach allows us to perform accurate forward uncertainty propagation consistent with the reported statistics.

97 MATHEMATICS AND COMPUTING↗