SEARCH · Engineering Papers

Results for “Machine Learning Algorithms”

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.

SourceResearch connectionFrom yearThrough year

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

At least 1,495 records · Page 83

All sources NASA NTRS DOE OSTI

DOE OSTI2021

Striving to translate shale physics across ten orders of magnitude: What have we learned?

Shales will play an important role in the successful transition of energy from fossil-based resources to renewables in the coming decades. Aside from being a significant source of low-carbon intensity fuels, like natural gas, they also serve as geologic seals of subsurface formations that may be used to isolate nuclear waste, sequester CO 2 , or store intermittent energy (e.g., solar hydrogen). Despite their importance, shales pose significant engineering and environmental challenges due to their nanoporous structure and extreme heterogeneity that spans at least ~10 orders of magnitude in spatial scale. Two challenges inhibit a system-level understanding: (1) the physics of fluid flow and phase behavior in shales are poorly understood due to the dominant molecular interactions between minerals and fluids under confinement, and (2) the apparent lack of scale separation that prevents a reliable (closed) description of the physics at any single scale of observation. In this review, we focus on the latter issue and discuss scale translation, which in its broadest sense is transforming data or simulations from one spatiotemporal scale to another. While effective scale translation is not exclusive to shales, but all geologic porous media, the need for it is especially acute in shales given their high degree of heterogeneity. Classical theories like homogenization, while indispensable, fail when scales are not separated. Other methods, like numerical upscaling, scale-translate in only one direction: small to large, but not the reverse, called downscaling. However, the confluence of advances in three areas are bringing challenging problems such as shales within reach: increased computational power and scalable algorithms; high-resolution imaging and multi-modal data acquisition; and machine learning to process massive amounts of data. While these advances equip geoscientists with a wide array of experimental and computational tools, no individual tool can probe the entire gamut of heterogeneity in shales. Their effective use, therefore, requires an ability to bridge between various data types obtained at different scales. The aim of this review is to present a coherent account of computational and experimental methods that may be used to achieve just that, i.e., to perform scale translation. We provide a broader definition of scale translation, one that transcends classical homogenization and upscaling methods, but is consistent with them and accommodates notions like downscaling and data translation. After a brief introduction to homogenization, we review hybrid methods, numerical upscaling and its recent extensions, multiscale computing, high-resolution imaging, and machine learning. We place particular emphasis on multiscale computing and propose an algorithmic framework to bridge between the pore (micro) and Darcy (macro) scales. Throughout the paper, we draw comparisons between the various methods and highlight their (often hidden) similarities, differences, benefits, and pitfalls. We finally conclude with two case studies on shales that exemplify some of the methods presented.

58 GEOSCIENCES↗

DOE OSTI2023

Classification of four-qubit entangled states via machine learning

We apply the support vector machine (SVM) algorithm to derive a set of entanglement witnesses (EW) to identify entanglement patterns in families of four-qubit states. The effectiveness of SVM for practical EW implementations stems from the coarse-grained description of families of equivalent entangled quantum states. The equivalence criteria in our work is based on the stochastic local operations and classical communication classification and the description of the four-qubit entangled Werner states. We numerically verify that the SVM approach provides an effective tool to address the entanglement witness problem when the coarse-grained description of a given family state is available. Here, we also discuss and demonstrate the efficiency of nonlinear kernel SVM methods as applied to four-qubit entangled state classification.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

DOE OSTI2019

Quantifying Climate and Catchment Control on Hydrological Drought in the Continental United States

The evolution of hydrological drought events is a result of complex (nonlinear) interactions between climate and catchment processes. To investigate such nonlinear relationship, we integrated a machine learning modeling framework based on the random forest (RF) algorithms with an interpretation framework to quantify the role of climate and catchment controls on hydrological drought. More particularly, our framework interprets a built RF machine-learning model to identify dominant variables and visualize their functional dependence and interaction effects on hydrological drought characteristics utilizing concepts of minimal depth, interactive depth, and partial dependence. We test our proposed modeling framework based on a set of 652 continental United States catchments with minimal human interference for a period of 1979–2010. Application of this framework indicated presence of three distinct drought regimes, which includes, Regime 1: droughts with longer duration, less frequent and lesser intensity; Regime 2: droughts with moderate duration, moderate frequency, and moderate intensity; and Regime 3: droughts with shorter duration, more frequent, and more intense. RF algorithm was able to accurately model the drought characteristics (intensity, duration, and number of events) for all the three drought regimes as a function of selected variables. It was observed that the type of dominant variables as well as their nonlinear functional relationship with hydrological droughts characteristics can vary between three selected regimes. Finally, our interpretation framework indicated that catchment characteristics have a significant role in controlling the hydrologic drought for catchments (regime 1), whereas both climate and catchment characteristics control hydrological drought in regimes 2 and 3.

54 ENVIRONMENTAL SCIENCES↗

DOE OSTI2018

Preprocessing Prediction of Advanced Algorithms for Medical Imaging

Advanced medical imaging algorithms (such as bone removal, vessel segmentation, or a lung nodule detection) can provide extremely valuable information to the radiologists, but they might sometimes be very time consuming. Being able to run the algorithms in advance can be a possible solution. However, we do not know which algorithm to run on a given dataset before it is actually used. It is possible to manually insert matching rules for preprocessing algorithms, but it requires high maintenance and does not work well in practice. This paper presents a dynamic machine learning solution for predicting which advanced visualization (AV) algorithm needs to be applied on a given series. The system gets a handful of free text DICOM tags as an input and builds a model in the clinical setting. It incorporates a Bag of Words (BOW) feature extractor and a Random Forest classifier. The approach was tested on two datasets from clinical sites which use different languages and varying scanner models. We show that even without feature extraction, sensitivity of above 90% can be reached on both of them. By using BOW feature extractor, precision and sensitivity can usually be further improved. Even on a noisy and highly unbalanced dataset, only around 100 samples were needed to reach sensitivity of above 80% and specificity of above 97%. We show how the solution can be part of a Smart Preprocessing mechanism in a viewing software. Using such a system will ultimately minimize the time to launch studies and improve radiologists reading time efficiency.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

DOE OSTI2023

The influence of physical and algorithmic factors on simulated far-field waveforms and source–time functions of underground explosions using unsupervised machine learning

SUMMARY Characterizing explosion sources and differentiating between earthquake and underground explosions using distributed seismic networks becomes non-trivial when explosions are detonated in cavities or heterogeneous ground material. Moreover, there is little understanding of how changes in subsurface physical properties affect the far-field waveforms we record and use to infer information about the source. Simulations of underground explosions and the resultant ground motions can be a powerful tool to systematically explore how different subsurface properties affect far-field waveform features, but there are added variables that arise from how we choose to model the explosions that can confound interpretation. To assess how both subsurface properties and algorithmic choices affect the seismic wavefield and the estimated source functions, we ran a series of 2-D axisymmetric non-linear numerical explosion experiments and wave propagation simulations that explore a wide array of parameters. We then inverted the synthetic far-field waveform data using a linear inversion scheme to estimate source–time functions (STFs) for each simulation case. We applied principal component analysis (PCA), an unsupervised machine learning method, to both the far-field waveforms and STFs to identify the most important factors that control variance in the waveform data and differences between cases. For the far-field waveforms, the largest variance occurs in the shallower radial receiver channels in the 0–50 Hz frequency band. For the STFs, both peak amplitude and rise times across different frequencies contribute to the variance. We find that the ground equation of state (i.e. lithology and rheology) and the explosion emplacement conditions (i.e. tamped versus cavity) have the greatest effect on the variance of the far-field waveforms and STFs, with the ground yield strength and fracture pressure being secondary factors. Differences in the PCA results between the far-field waveforms and STFs could possibly be due to near-field non-linearities of the source that are not accounted for in the estimation of STFs and could be associated with yield strength, fracture pressure, cavity radius and cavity shape parameters. Other algorithmic parameters are found to be less important and cause less variance in both the far-field waveforms and STFs, meaning algorithmic choices in how we model explosions are less important, which is encouraging for the further use of explosion simulations to study how physical Earth properties affect seismic waveform features and estimated STFs.

58 GEOSCIENCES↗

NASA NTRS1991

Paradigms for machine learning

Five paradigms are described for machine learning: connectionist (neural network) methods, genetic algorithms and classifier systems, empirical methods for inducing rules and decision trees, analytic learning methods, and case-based approaches. Some dimensions are considered along with these paradigms vary in their approach to learning, and the basic methods are reviewed that are used within each framework, together with open research issues. It is argued that the similarities among the paradigms are more important than their differences, and that future work should attempt to bridge the existing boundaries. Finally, some recent developments in the field of machine learning are discussed, and their impact on both research and applications is examined.

Schlimmer, Jeffrey C.↗

DOE OSTI2020

Benchmarking materials property prediction methods: the Matbench test set and Automatminer reference algorithm

Abstract We present a benchmark test suite and an automated machine learning procedure for evaluating supervised machine learning (ML) models for predicting properties of inorganic bulk materials. The test suite, Matbench, is a set of 13 ML tasks that range in size from 312 to 132k samples and contain data from 10 density functional theory-derived and experimental sources. Tasks include predicting optical, thermal, electronic, thermodynamic, tensile, and elastic properties given a material’s composition and/or crystal structure. The reference algorithm, Automatminer, is a highly-extensible, fully automated ML pipeline for predicting materials properties from materials primitives (such as composition and crystal structure) without user intervention or hyperparameter tuning. We test Automatminer on the Matbench test suite and compare its predictive power with state-of-the-art crystal graph neural networks and a traditional descriptor-based Random Forest model. We find Automatminer achieves the best performance on 8 of 13 tasks in the benchmark. We also show our test suite is capable of exposing predictive advantages of each algorithm—namely, that crystal graph methods appear to outperform traditional machine learning methods given ~10 4 or greater data points. We encourage evaluating materials ML algorithms on the Matbench benchmark and comparing them against the latest version of Automatminer.

36 MATERIALS SCIENCE↗

DOE OSTI2020

Development of Prognostic Models Using Plant Asset Data

The recent growth of machine learning and artificial intelligence technologies provides opportunities for leveraging data-driven algorithms to address the problems of diagnostics and prognostics in the nuclear power industry. The use of machine learning and other statistical methods as prognostic models is of particular interest in the nuclear industry to accurately predict future equipment or plant state given a set of measurements. Such predictive capability will enable predictive assessment of component condition and remaining life and allow for condition-based predictive maintenance. The resulting optimization of maintenance scheduling and reduction in unnecessary maintenance activities will lower overall maintenance costs and improve the economics of nuclear power. This report discusses the various aspects of data processing and model development that are likely to influence the performance of prognostic models. Data from a boiling-water reactor was used to evaluate several prognostic models to identify key considerations for developing such models to predict data-driven plant state and equipment degradation condition. Preliminary results indicate the need for data sets that are relevant to the problem at hand and contain signatures that may be correlated to the prediction problem. Assuming such data exist, development of prognostic models using data-driven methods requires an understanding of the various sources of influence on the prediction accuracy (such as the model architecture, data preprocessing approaches, and potentially external factors influencing the equipment or plant system under assessment). Ongoing research is evaluating these factors in greater detail and examining techniques for calculating prediction uncertainty bounds.

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗

DOE OSTI2022

MIDAS: Modeling Individual Differences using Advanced Statistics

This research explores novel methods for extracting relevant information from EEG data to characterize individual differences in cognitive processing. Our approach combines expertise in machine learning, statistics, and cognitive science, advancing the state-of-the art in all three domains. Specifically, by using cognitive science expertise to interpret results and inform algorithm development, we have developed a generalizable and interpretable machine learning method that can accurately predict individual differences in cognition. The output of the machine learning method revealed surprising features of the EEG data that, when interpreted by the cognitive science experts, provided novel insights to the underlying cognitive task. Additionally, the outputs of the statistical methods show promise as a principled approach to quickly find regions within the EEG data where individual differences lie, thereby supporting cognitive science analysis and informing machine learning models. This work lays methodological ground work for applying the large body of cognitive science literature on individual differences to high consequence mission applications.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2022

Layer-wise Imaging Dataset from Powder Bed Additive Manufacturing Processes for Machine Learning Applications (Peregrine v2022-10)

This release consists of six datasets which together include multi-modal layer-wise powder bed images from two different powder bed printing technologies. These datasets are designed primarily to facilitate the development and testing of new computer vision and machine learning based anomaly and defect detection algorithms. The authors provide both training data with corresponding ground truth pixel masks and evaluation data with corresponding baseline prediction pixel masks made by a trained neural network. The laser powder bed fusion (L-PBF) datasets are sourced from EOS M290 and AddUp FormUp 350 printers and the binder jet (BJ) dataset is sourced from an ExOne M-Flex printer. The materials represented in these datasets include 17-4 PH Stainless Steel, DMREF, Inconel 718, Maraging Steel, and H13 Steel. The sensor imaging modalities represented include visible-light (VL), temporally-integrated (i.e., long duration exposure) near-infrared (TI-NIR), and wide-band infrared (IR).

36 MATERIALS SCIENCE↗

DOE OSTI2023

Using angular two-point correlations to self-calibrate the photometric redshift distributions of DECaLS DR9

Calibrating the redshift distributions of photometric galaxy samples is essential in weak lensing studies. The self-calibration method combines angular auto- and cross-correlations between galaxies in multiple photometric redshift (photo-z) bins to reconstruct the scattering rates matrix between redshift bins. In this paper, we test a recently proposed self-calibration algorithm using the DECaLS Data Release 9 and investigate to what extent the scattering rates are determined. We first mitigate the spurious angular correlations due to imaging systematics by a machine learning based method. We then improve the algorithm for χ 2 minimization and error estimation. Finally, we solve for the scattering matrices, carry out a series of consistency tests, and find reasonable agreements: (1) finer photo-z bins return a high-resolution scattering matrix, and it is broadly consistent with the low-resolution matrix from wider bins; (2) the scattering matrix from the Northern Galactic Cap is almost identical to that from Southern Galactic Cap; (3) the scattering matrices are in reasonable agreement with those constructed from the power spectrum and the weighted spectroscopic subsample. We also evaluate the impact of cosmic magnification. Although it changes little the diagonal elements of the scattering matrix, it affects the off-diagonals significantly. The scattering matrix also shows some dependence on scale cut of input correlations, which may be related to a known numerical degeneracy between certain scattering pairs. In conclusion, this work demonstrates the feasibility of the self-calibration method in real data and provides a practical alternative to calibrate the redshift distributions of photometric samples.

79 ASTRONOMY AND ASTROPHYSICS↗

DOE OSTI2018

Reduced-order modeling through machine learning and graph-theoretic approaches for brittle fracture applications

Typically, thousands of computationally expensive micro-scale simulations of brittle crack propagation are needed to upscale lower length scale phenomena to the macro-continuum scale. Running such a large number of crack propagation simulations presents a significant computational challenge, making reduced-order models (ROMs) attractive for this task. Here, the ultimate goal of this research is to develop ROMs that have sufficient accuracy and low computational cost so that these upscaling simulations can be readily performed. However, constructing ROMs for these complex simulations presents its own challenge. Here, we present and compare four different approaches for reduced-order modeling of brittle crack propagation in geomaterials. These methods rely on machine learning (ML) and graph-theoretic algorithms to approximate key aspects of the brittle crack problem. These methods also incorporate different physics-based assumptions in order to reduce the training requirements while maintaining accurate physics as much as possible. Results from the ROMs are directly compared against a high-fidelity model of brittle crack propagation. Further, the strengths and weaknesses of the ROMs are discussed, and we conclude that combining smart physics-informed feature engineering with highly trainable ML models provides the best performance. The ROMs considered here have computational costs that are orders-of-magnitude less than the cost associated with high-fidelity physical models while maintaining good accuracy.

36 MATERIALS SCIENCE↗

DOE OSTI2022

Selecting Post-Processing Schemes for Accurate Detection of Small Objects in Low-Resolution Wide-Area Aerial Imagery

In low-resolution wide-area aerial imagery, object detection algorithms are categorized as feature extraction and machine learning approaches, where the former often requires a post-processing scheme to reduce false detections and the latter demands multi-stage learning followed by post-processing. In this paper, we present an approach on how to select post-processing schemes for aerial object detection. We evaluated combinations of each of ten vehicle detection algorithms with any of seven post-processing schemes, where the best three schemes for each algorithm were determined using average F-score metric. The performance improvement is quantified using basic information retrieval metrics as well as the classification of events, activities and relationships (CLEAR) metrics. We also implemented a two-stage learning algorithm using a hundred-layer densely connected convolutional neural network for small object detection and evaluated its degree of improvement when combined with the various post-processing schemes. The highest average F-scores after post-processing are 0.902, 0.704 and 0.891 for the Tucson, Phoenix and online VEDAI datasets, respectively. The combined results prove that our enhanced three-stage post-processing scheme achieves a mean average precision (mAP) of 63.9% for feature extraction methods and 82.8% for the machine learning approach.

54 ENVIRONMENTAL SCIENCES↗

DOE OSTI2021

Lithium-ion battery physics and statistics-based state of health model

A pseudo-2d model using COMSOL Multiphysics® software is developed to simulate performance and performance degradation of Li-ion batteries consisting of layered and olivine cathodes with graphite anode when subjected to peak shaving grid service. Multiple degradation pathways are considered, including solid electrolyte interphase (SEI) formation and breakdown at the anode, cathode dissolution and its synergistic effect on SEI formation at the anode. The model is validated by simulating commercial cylindrical cell performance. A global model is developed to simulate performance across all chemistries, along with individual chemistry models using global model parameters as initial values. There is good agreement between these models for various optimization parameters such as SEI equilibrium potential, cathode dissolution exchange current density, solvent diffusivity in the SEI and SEI ionic conductivity. To circumvent time constraints related to the COMSOL model, a 0d global model is developed which fits data well and provides more clarity on differences in cathode dissolution exchange current density. Again, good agreement for various optimization parameters is obtained among the COMSOL global & individual chemistry models and the 0-d model. The lessons learned from the physics-based model is used to develop a top down statistics-based model using current, voltage and anode volumetric change per mole lithium intercalated, along with their interactions as degradation predictors. This model predicts out of sample degradation for multiple grid services and electric vehicle drive cycle with high accuracy and provides the pathway to develop an efficient battery management system combining machine learning and findings from physics-based computationally intensive algorithms.

Crawford, Aladsair J.↗

DOE OSTI2022

Cyber risk assessment and investment optimization using game theory and ML-based anomaly detection and mitigation for wide-area control in smart grids

The electric power grid is increasingly becoming susceptible to cyber attacks that exploit vulnerabilities in the smart grid control, information, and physical layers. Successful cyber attacks can have catastrophic impacts on the social and economic well-being of any nation all over the globe. It has, thus, become imperative to secure the smart grid against such adversarial actions to ensure stable, secure, and reliable operation of the grid. The existing research and industry practices prove to be inadequate in terms of providing pragmatic and effective defense methodologies and measures for long-term cybersecurity planning and real-time cybersecurity for grid operation. For example, existing works lack models that incorporate uncertain behavior of cyber-attackers and pragmatic defense measures for cyber risk assessment and cybersecurity investment optimization which often provide unreliable and strictly qualitative solutions to these problems. At the same time, with the growing number of cyber incidents in the grid, there still exists a need to develop attack-resilient algorithms for wide-area monitoring, protection, and control (WAMPAC) applications like the wide-area voltage control systems (WAVCS) for Flexible AC Transmissions Systems (FACTS) that lack in scalable and feasible solutions from the cybersecurity perspective. This dissertation proposes novel models and methodologies for: (1) Cybersecurity planning, and (2) Cybersecurity for system operation. The cybersecurity planning is achieved through cyber risk assessment and cybersecurity resource investment optimization for long-term cybersecurity of the grid using game theory and attack-defense trees. Cybersecurity for system operation consists of development of cyber anomaly detection and mitigation algorithms for flexible AC transmission system (FACTS) controller-based wide-area voltage control systems (WAVCS) using machine learning (ML), and software defined networking-based moving target defense network routing for achieving real-time cyber-physical security for grid operations. This is followed by hardware-in-the-loop (HIL) implementation and evaluation of these attack prevention, detection, and mitigation algorithms and methodologies showcasing their feasibility in a close to real-world environment. For cybersecurity planning, a novel approach involving a combination of game theory and attack defense trees (ADT) for optimal cybersecurity resource allocation in the smart grid is proposed. This methodology involves modeling of the cyber-physical smart grid substations as ADTs, defining attacker costs, defense costs, and attack probabilities for attack access points. Using game theoretical formulation, optimal defense strategies for the defender of the system to invest cybersecurity resources in the grid are obtained. Additionally, a game-theoretic framework is developed for quantitative cyber-physical risk assessment of the grid under a dynamically changing cyber threat space and uncertain behavior of cyber attackers which is further used to optimize investments in the smart grid's cybersecurity resources. The attacker, defender, and the smart grid system are modeled while incorporating attacker-stochasticity and federal guidelines for smart grid cybersecurity. This allows quantification of threat, vulnerabilities, and attack impact of the grid for quantitative risk assessment. The defender's budget to invest in the security resources in the grid is optimized based on the strategies leading to minimum system risk. The evaluation of the proposed solutions highlight the feasibility for practical implementation of these methodologies and algorithms in the smart grid, while taking the federal requirements and guidelines for smart grid security into consideration. For achieving cybersecurity for system operation, attack prevention, detection, and mitigation algorithms and methodologies are developed specifically for FACTS-based WAVCS. Anomaly detection and mitigation in the WAVCS are achieved using algorithms based on machine learning which involves offline training and testing of ML models with CPS datasets incorporating physics-based features that allow accurate distinction between system faults and cyber attacks. For attack prevention, a methodology based on software defined network (SDN)-based moving target defense (MTD) network routing is proposed that enables prevention of Denial of Service (DoS) type attacks on the smart grid communication system. Subsequently, these methodologies and algorithms are implemented and evaluated on an HIL testbed that allows for real-time attack prevention, detection, and mitigation of emulated cyber attacks on the WAVCS in a close to real-world environment. The results show highly accurate and efficient performance of the implemented algorithms and methodologies with the smart grid system operating within the NERC's system operation limits even in the presence of DoS and data integrity cyber attacks. This work opens up future research opportunities in other directions such as (1) Expanding cybersecurity planning methodologies to real-time cyber contingency analysis with different game formulations; and (2) Applying the cybersecurity for system operation algorithms to broader categories of wide-area control applications.

24 POWER TRANSMISSION AND DISTRIBUTION↗

NASA NTRS1990

The mGA1.0: A common LISP implementation of a messy genetic algorithm

Genetic algorithms (GAs) are finding increased application in difficult search, optimization, and machine learning problems in science and engineering. Increasing demands are being placed on algorithm performance, and the remaining challenges of genetic algorithm theory and practice are becoming increasingly unavoidable. Perhaps the most difficult of these challenges is the so-called linkage problem. Messy GAs were created to overcome the linkage problem of simple genetic algorithms by combining variable-length strings, gene expression, messy operators, and a nonhomogeneous phasing of evolutionary processing. Results on a number of difficult deceptive test functions are encouraging with the mGA always finding global optima in a polynomial number of function evaluations. Theoretical and empirical studies are continuing, and a first version of a messy GA is ready for testing by others. A Common LISP implementation called mGA1.0 is documented and related to the basic principles and operators developed by Goldberg et. al. (1989, 1990). Although the code was prepared with care, it is not a general-purpose code, only a research version. Important data structures and global variations are described. Thereafter brief function descriptions are given, and sample input data are presented together with sample program output. A source listing with comments is also included.

Goldberg, David E.↗

DOE OSTI2024

AXI4MLIR: User-Driven Automatic Host Code Generation for Custom AXI-Based Accelerators

Tensor algebra operations represent an important class of algorithms used across many applications, including machine learning, scientific computing, and data analytics. As a result, the efficient generation of custom accelerators for tensor operations has received increased attention. Previous efforts have produced automated tools enabling users to prototype and explore optimized accelerators. However, little effort has been focused on the host-accelerator interaction in these tools. Efficient use of hardware accelerators requires knowledge about the accelerator's capabilities (operations, data formats, and opcode support), the host CPU microarchitecture (e.g., memory hierarchy), the host-accelerator interface, and the application's features (which code regions should be mapped onto an accelerator). Manually rewriting the original applications to facilitate improved custom accelerator mapping is an error-prone and time-consuming endeavor. To cope with this, we propose AXI4MLIR, a new framework to automatically generate and optimize the communication between the host CPU and arbitrary accelerators that implement linear algebra algorithms. AXI4MLIR extends the MLIR compiler framework to automatically generate efficient host-accelerator driver code for accelerators with AXI-based interfaces. Our compiler extensions enable automatic driver code generation while carefully considering the host's memory hierarchy and target accelerator features. To demonstrate the flexibility and utility of AXI4MLIR, we test it with diverse use cases that include different types of accelerators, tiling scenarios, and dataflow schemes. We compare our experimental results to manual implementations of host-accelerator driver code and find that our approach can reduce CPU cache references by 56% and deliver up to a 1.65x speedup.

Bohm Agostini, Nicolas↗

DOE OSTI2025

Neutrino interaction vertex reconstruction in DUNE with Pandora deep learning

The Pandora Software Development Kit and algorithm libraries perform reconstruction of neutrino interactions in liquid argon time projection chamber detectors. Pandora is the primary event reconstruction software used at the Deep Underground Neutrino Experiment, which will operate four large-scale liquid argon time projection chambers at the far detector site in South Dakota, producing high-resolution images of charged particles emerging from neutrino interactions. While these high-resolution images provide excellent opportunities for physics, the complex topologies require sophisticated pattern recognition capabilities to interpret signals from the detectors as physically meaningful objects that form the inputs to physics analyses. A critical component is the identification of the neutrino interaction vertex. Subsequent reconstruction algorithms use this location to identify the individual primary particles and ensure they each result in a separate reconstructed particle. A new vertex-finding procedure described in this article integrates a U-ResNet neural network performing hit-level classification into the multi-algorithm approach used by Pandora to identify the neutrino interaction vertex. The machine learning solution is seamlessly integrated into a chain of pattern-recognition algorithms. The technique substantially outperforms the previous BDT-based solution, with a more than 20% increase in the efficiency of sub-1 cm vertex reconstruction across all neutrino flavours.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗