Search NASA⌕ Search

SEARCH · Search NASA

Results for “acquisition function”

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 55 records · Page 3

Sequential Selection for Minimizing the Variance with Application to Crystallization Experiments

For many crystal-based products (e.g., pharmaceuticals, energy storage), the size uniformity is not only a key quality attribute, but sometimes also an indicator of other attributes such as solid purity. This article proposes a sequential selection approach to find a proper experimental setting that leads to high uniformity, or equivalently, small variance for crystal sizes, from the advanced slug flow reaction crystallization process of a model crystal, called manganese oxalate hydrate. The proposed sequential selection approach contains a Bayesian adaptive method to incorporate new uniformity measurements in each step and two design acquisition functions to improve the selection of the most promising experimental setting in terms of minimizing the variance. We study the performance of the proposed approach through multiple synthetic numerical studies, as well as a case study based on data from slug flow crystallization experiments. Throughout these studies, the proposed approach shows competitive performance in identifying the best experimental setting.

Expected improvement↗

Automated experimental design of safe rampdowns via probabilistic machine learning

Abstract Typically the rampdown phase of a shot consists of a decrease in current and injected power and optionally a change in shape, but there is considerable flexibility in the rate, sequencing, and duration of these changes. On the next generation of tokamaks it is essential that this is done safely as the device could be damaged by the stored thermal and electromagnetic energy present in the plasma. This works presents a procedure for automatically choosing experimental rampdown designs to rapidly converge to an effective rampdown trajectory. This procedure uses probabilistic machine learning methods paired with acquisition functions taken from Bayesian optimization. In a set of 2022 experiments at DIII-D, the rampdown designs produced by our method maintained plasma control down to substantially lower current and energy levels than are typically observed. The actions predicted by the model significantly improved as the model was able to explore over the course of the experimental campaign.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Human-in-the-Loop: The Future of Machine Learning in Automated Electron Microscopy

Machine learning (ML) methods are progressively gaining acceptance in the electron microscopy community for de-noising, semantic segmentation, and dimensionality reduction of data post-acquisition. The introduction of the application programming interfaces (APIs) by major instrument manufacturers now allows the deployment of ML workflows in microscopes, not only for data analytics but also for real-time decision-making and feedback for microscope operation. However, the number of use cases for real-time ML remains remarkably small. Furthermore, we discuss some considerations in designing ML-based active experiments and pose that the likely strategy for the next several years will be human-in-the-loop automated experiments (hAE). In this paradigm, the ML learning agent directly controls beam position and image and spectroscopy acquisition functions, and a human operator monitors experiment progression in real and feature space of the system and tunes the policies of the ML agent to steer the experiment toward specific objectives.

47 OTHER INSTRUMENTATION↗

PABO: Pseudo Agent-Based Multi-Objective Bayesian Hyperparameter Optimization for Efficient Neural Accelerator Design

The ever increasing computational cost of Deep Neural Networks (DNN) and the demand for energy efficient hardware for DNN acceleration has made accuracy and hardware cost co-optimization for DNNs tremendously important, especially for edge devices. Owing to the large parameter space and cost of evaluating each parameter in the search space, manually tuning of DNN hyperparameters is impractical. Automatic joint DNN and hardware hyperparameter optimization is indispensable for such problems. Bayesian optimization-based approaches have shown promising results for hyperparameter optimization of DNNs. However, most of these techniques have been developed without considering the underlying hardware, thereby leading to inefficient designs. Further, the few works that perform joint optimization are not generalizable and mainly focus on CMOS-based architectures. In this work, we present a novel pseudo agent-based multiobjective hyperparameter optimization (PABO) for maximizing the DNN performance while obtaining low hardware cost. Compared to the existing methods, our work poses a theoretically different approach for joint optimization of accuracy and hardware cost and focuses on memristive crossbar based accelerators. PABO uses a supervisor agent to establish connections between the posterior Gaussian distribution models of network accuracy and hardware cost requirements. The agent reduces the mathematical complexity of the co-optimization problem by removing unnecessary computations and updates of acquisition functions, thereby achieving significant speed-ups for the optimization procedure. PABO outputs a Pareto frontier that underscores the trade-offs between designing high-accuracy and hardware efficiency. Our results demonstrate a superior performance compared to the state-of-the-art methods both in terms of accuracy and computational speed (~100x speed up).

Parsa, Maryam↗

GP Cosmology Surrogate v1.0

GP Cosmology Surrogate is a Python library for building and training a generalized multi-output Gaussian process (GP) framework of @takhtaganov2021cosmic. In this approach, the surrogate is constructed sequentially, guided by a Bayesian optimization acquisition function that targets reduction of emulation error in the regions most consistent with the observational data. This adaptive design concentrates computational resources where they have the greatest impact on inference accuracy. The library supports efficient training for separable GP kernels, which allows the use of Kronecker algebra to handle high-dimensional input spaces and large numbers of correlated outputs. This makes it well suited for applications such as modeling cosmological power spectra, large-scale physical simulations, and multi-output hyperparameter tuning. By combining scalable multi-output GP modeling with data-driven adaptive sampling, GPsurrogate enables parameter inference and optimization with substantially fewer simulations than conventional space-filling designs.

Lukic, Zarija [Lawrence Berkeley National Laborato↗

Designing a Framework for Solving Multiobjective Simulation Optimization Problems

Multiobjective simulation optimization (MOSO) problems are optimization problems with multiple conflicting objectives, where evaluation of at least one of the objectives depends on a black-box numerical code or real-world experiment, which we refer to as a simulation. Whereas an extensive body of research is dedicated to developing new algorithms and methods for solving these and related problems, it is challenging and time-consuming to integrate these techniques into real-world production-ready solvers. This is partly because of the diversity and complexity of modern state-of-the-art MOSO algorithms and methods and partly because of the complexity and specificity of many real-world problems and their corresponding computing environments. The complexity of this problem is only compounded when introducing potentially complex and/or domain-specific surrogate-modeling techniques, problem formulations, design spaces, and data acquisition functions. Here, this paper carefully surveys the current state of the art in MOSO algorithms, techniques, and solvers, as well as problem types and computational environments where MOSO is commonly applied. We then present several key challenges in the design of a parallel multiobjective simulation optimization framework (ParMOO) and how they have been addressed. Finally, we provide two case studies demonstrating how customized ParMOO solvers can be quickly built and deployed to solve real-world MOSO problems.

engineering design optimization↗

Integrating adaptive learning with post hoc model explanation and symbolic regression to build interpretable surrogate models

Abstract We develop a materials informatics workflow to build an interpretable surrogate model for micromagnetic simulations. Our goal is to predict the energy barrier of a moving isolated skyrmion in rare-earth-free $$\hbox {Mn}_4$$ Mn 4 N. Our approach integrates adaptive learning with post hoc model explanation and symbolic regression methods. We discuss an unexplored acquisition function (information condensing active learning) within the adaptive learning loop and compare it with the known standard deviation function for efficient navigation of the search space. Model-agnostic post hoc explanation techniques then uncover trends learned by the trained model, which we then leverage to constrain the expressions used for symbolic regression. Graphical abstract

Biswas, Ankita↗

Active Learning for Language Modeling

Foreign disinformation campaigns undermine national security. Various supervised language modeling techniques in NLP can help to understand and dismantle these campaigns, but they rely heavily on large, labeled (often by humans) datasets. This work provides a solution to this problem in the form of an active learning (AL) framework, which is used to generate labeled datasets and leverage human input for detecting disinformation. The developed AL framework utilizes task adaptive pretraining to fully leverage the unlabeled data and boost the performance of the classifier used for labeling. A disinformation rhetoric metric was developed to measure the presence of common rhetorical techniques used in text that are meant to deceive, for both the classifier and human to use in the task of identifying disinformation. This metric was combined with an uncertainty criterion to create a hybrid acquisition method for AL, and this hybrid method was tested alongside other acquisition functions. A sophisticated and robust stopping strategy was developed to signal the AL process should terminate, saving human time from being wasted on iterations that would not significantly benefit classifier performance.

99 GENERAL AND MISCELLANEOUS↗

Deep Active Learning for Solvability Prediction in Power Systems

Traditional methods for solvability region analysis can only have inner approximations with inconclusive conservatism and handle limited types of power flow models. In this letter, we propose a deep active learning framework for solvability prediction in power systems. Compared with passive learning where the training is performed after all instances are labeled, active learning selects most informative instances to be labeled and therefore significantly reduces the size of the labeled dataset for training. In the active learning framework, the acquisition functions, which correspond to different sampling strategies, are defined in terms of the on-the-fly posterior probability from the classifier. First, the IEEE 39-bus system is employed to validate the proposed framework, where a two-dimensional case is illustrated to visualize the effectiveness of the sampling method followed by the high-dimensional numerical experiments. Then, the Northeast Power Coordinating Council (NPCC) 140-bus system is used to validate the performance on large-scale power systems.

42 ENGINEERING↗

The Kalai-Smorodinsky solution for many-objective Bayesian optimization

An ongoing aim of research in multiobjective Bayesian optimization is to extend its applicability to a large number of objectives. While coping with a limited budget of evaluations, recovering the set of optimal compromise solutions generally requires numerous observations and is less interpretable since this set tends to grow larger with the number of objectives. We thus propose to focus on a specific solution originating from game theory, the Kalai-Smorodinsky solution, which possesses attractive properties. In particular, it ensures equal marginal gains over all objectives. We further make it insensitive to a monotonic transformation of the objectives by considering the objectives in the copula space. A novel tailored algorithm is proposed to search for the solution, in the form of a Bayesian optimization algorithm: sequential sampling decisions are made based on acquisition functions that derive from an instrumental Gaussian process prior. Our approach is tested on four problems with respectively four, six, eight, and nine objectives. The method is available in the R package GPGame available on CRAN at https://cran.r-project.org/package=GPGame.

Game theory↗

MOOSE ProbML: Parallelizable Probabilistic Machine Learning and Uncertainty Quantification Capabilities

The Multiphysics Object Oriented Simulation Environment (MOOSE) is a widely used open- source finite element software for performing multiphysics multiscale simulations in a massively parallel fashion. Recently, the computational team at Idaho National Laboratory (INL) has implemented Probabilistic Machine Learning (ProbML) capabilities in MOOSE—in a parallelized fashion—and enable active learning with large-scale computational models for tasks such as surrogate model development, scale bridging, forward/inverse uncertainty quantification (UQ), Bayesian optimization, etc. This presentation summarizes these developments in MOOSE along with demonstrations on several real applications relevant to nuclear energy. At the fundamental level, samplers like Monte Carlo/Latin Hypercube, variance reduction, parallelized Markov Chain Monte Carlo (MCMC) support uncertainty propagation in both forward and inverse settings. These samplers can be integrated with the Gaussian processes (GP) suite in MOOSE, which offer several variants like scalar GPs, multi-output GPs, and deep GPs, to enable active learning. These GPs can be tuned using gradient-based optimization methods like Adam and its variants or gradient-free methods like the elliptical slice sampler (a variant of MCMC adept under Gaussian settings) for more complex covariance kernels or likelihoods whose gradient computations can be cumbersome. A variety of batch acquisition functions permit parallelized evaluation of the computational model and support different learning objectives with high efficiency like Bayesian inference, global surrogate development, optimization, etc. Furthermore, libtorch integration supports training, evaluation, and re-training of neural networks and other complex machine learning models in active learning settings. The impacts of these developments are shown on several real applications: (1) nuclear fuel inverse UQ and model inadequacy assessment using the Kennedy O’Hagan framework; (2) uncertainty aware surrogate modeling for additive manufacturing to predict field quantities; (3) nuclear reactor rare events analysis; and (4) complex fluid flow prediction using a global surrogate with quantified prediction uncertainty. Finally, the outlook of MOOSE ProbML is discussed for both outer-loop and inner-loop computations in the broad view to accelerate fuels and materials qualification, address gaps in knowledge and data, and assess new reactor/fuel systems.

11 - NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗

Bayesian Optimization for Reactor Design Optimization

This study present a test case in which the Bayesian Optimization method is applied to a simulation-based reactor core design optimization problem. The test case aims to showcase the potential of an automated design optimization algorithm for reactor designs by streamlining the reactor core design workflow, given the high computational cost of simulations. The contributions of this work are threefold. First, the existing HTGR model is converted into a simulation-based design optimization test case by developing a pipeline that enables modification of key design parameters and evaluates design performance based on simulation outputs. Second, Bayesian Optimization is implemented and adapted to demonstrate the feasibility of automatic design optimization for nuclear reactor core. Proposed approach leverages Gaussian Process models to characterize the relationship between design variables and performance metrics, while incorporating novel acquisition functions that balance exploration of the design space with exploitation of promising configurations. This implementation lays the foundation for the future developments of reactor design optimization algorithms.

22 - GENERAL STUDIES OF NUCLEAR REACTORS↗

Physical and Functional Constraints on Viable Belowground Acquisition Strategies

Since their emergence onto land, terrestrial plants have developed diverse strategies to acquire soil resources. However, we lack a framework that adequately captures how these strategies vary among species. Observations from around the world now allow us to quantify the variation observed in commonly-measured fine-root traits but it is unclear how root traits are interrelated and whether they fall along an “economic” spectrum of acquisitive to conservative strategies. We assessed root trait variation and mycorrhizal colonization rates by leveraging the largest global database of fine-root traits (the Fine-Root Ecology Database; FRED). We also developed a heuristic model to explore the role of mycorrhizal fungi in defining belowground exploration efficiency across a gradient of thin- to thick-diameter roots. In support of the expectations of the “root economic spectrum,” we found that root diameter was negatively related to specific root length (Pearson’s r=-0.76). However, we found an unexpected negative relationship between root diameter and root tissue density (Pearson’s r = -0.40), and we further observed that root nitrogen content was largely unrelated to other economic traits. Mycorrhizal colonization was most closely associated with root diameter (Pearson’s r = 0.62) and was unrelated to root tissue density and root nitrogen. The heuristic model demonstrated that while thinner roots have inherently greater capacity to encounter soil resources based on higher surface area per unit mass, the potential for increased associations with mycorrhizal fungi in thicker roots, combined with greater hyphal growth, can result in equally acquisitive strategies for both thin- and thick roots. Taken together, our assessments of root trait variation, trade-offs with mycorrhizal fungi, and broader connections to root longevity allowed us to propose a series of fundamental constraints on belowground resource acquisition strategies. Physical tradeoffs based on root construction (i.e., economic traits) and functional limitations related to the capacity of a root to encounter and acquire soil resources combine to limit the two-dimensional belowground trait space. Within this trait space there remains a diversity of additional variation in root traits that facilitates a wide range of belowground resource acquisition strategies.

59 BASIC BIOLOGICAL SCIENCES↗

Mortality correlates with tree functional traits across a wood density gradient in the Central Amazon

Introduction: Understanding the mechanisms of tree mortality in tropical ecosystems remains challenging, in part due to the high diversity of tree species and the inherently stochastic nature of mortality. Plant functional traits offer a mechanistic link between plant physiology and performance, yet their ability to predict growth and mortality remains poorly understood. Given recent increases in tree mortality rates in the Amazon forest following extreme drought and wind events, we tested if lower wood density and acquisitive plant functional traits were associated with increased growth and mortality for common co-occurring trees in the Central Amazon. Methods: Seventeen trees of different species with similar sizes but a range in wood density (WD) and wood traits were felled, then assessed for 27 different individual functional parameters, including whole tree architecture, stem xylem anatomical and hydraulic traits and leaf traits. Traits of the individual trees were related to stand-level growth and mortality rates collected periodically over 30 years from nearby permanent inventory plots. Results: Higher wood density was associated with smaller leaf size, lower foliar base cations, lower stem water content and sapwood fraction, in agreement with the fast-slow plant economics spectrum. Lower wood density was associated with more acquisitive characteristics with greater hydraulic capacity and foliar nutrient concentrations, correlating with greater growth and mortality rates. Discussion: Our results show that lower wood density is part of a coordinated suite of traits linked to high resource acquisition, fast growth, and increased mortality risk, providing a functional framework for predicting species performance and forest vulnerability under future climate stress.

demographics↗

Tree architectural characteristics and stem and leaf functional traits for 17 individuals in the Central Amazon

Given recent increases in tree mortality rates in the Amazon forest following extreme drought and wind events, we tested if lower wood density and acquisitive plant functional traits were associated with increased growth and mortality for common co-occurring trees in the Central Amazon. Research was conducted at the ZF2 Research Station located north or Manaus, Brazil, managed by the Instituto Nacional de Pesquisas da Amazônia (INPA). Seventeen trees of different species with similar sizes but a range in wood density (WD) and wood traits were felled, then assessed for 27 different individual functional parameters, including whole tree architecture, stem xylem anatomical and hydraulic traits and leaf traits. Wood logs were collected at DBH, 50% stem length and at 100% stem length (at the base of the canopy). For wood anatomy samples, n=3-6 subsamples from each height. For leaf samples, 30 leaves were collected from the upper sunlit canopy. The methodology is detailed in the accompanying manuscript. The trait data are summarized in this file: "Trait_Summary.CSV". Summary Trait code abbreviations and units are described in this file: "Sample_Info_Traits_Summary.CSV". Stem traits measured along the bole from the base of the tree (DBH, diameter breast height), mid-stem, and base of the canopy are described in these files: "Sapwood_Area_height.CSV"; "Species_Info_height.CSV"; "Sample_Info_height.CSV"

54 ENVIRONMENTAL SCIENCES↗

Proteomic profiles of five strains of oxygenic photosynthetic cyanobacteria of the genus Cyanothece

Members of the cyanobacterial genus Cyanothece exhibit considerable variation in physiological and biochemical characteristics. The comparative assessment of the genomes and the proteomes has the potential to provide insights on differences among Cyanothece strains. By applying Sequedex (http://sequedex.lanl.gov), an annotationindependent method for ascribing gene functions, we confirmed significant speciesspecific differences of functional genes in different Cyanothece strains, particularly in Cyanothece PCC7425. Using a shotgun proteomics approach based on prefractionation and tandem mass spectrometry, we detected ~28-48% of the theoretical Cyanothece proteome depending on the strain. The expression of a total of 642 orthologous proteins was observed in all five Cyanothece strains. These shared orthologous proteins showed considerable correlations in their protein abundances across different Cyanothece strains. Functional classification indicated that the majority of proteins involved in central metabolic functions such as amino acid, carbohydrate, protein and RNA metabolism, photosynthesis, respiration and stress responses were observed to a greater extent in the core proteome, whereas proteins involved in membrane transport, iron acquisition, regulatory functions, flagellar motility and chemotaxis were observed to a greater extent in the unique proteome. Considerable differences were evident across different Cyanothece strains. Notably, the analysis of Cyanothece PCC7425, which showed the highest number of unique proteins (682),

Environmental Molecular Sciences Laboratory↗

TH-AB-209-04: 3D Light Sheet Luminescence Imaging with Cherenkov Radiation

Purpose: To recover a three-dimensional density distribution of luminescent molecular probes located several centimeters deep within a highly scattering tissue. Methods: We developed a novel sheet beam Cherenkov-excited luminescence scanned imaging (CELSI) methodology. The sample was irradiated by a horizontally oriented, vertically scanned 6 MV X-ray sheet beam (200mm × 5mm, 0.2mm vertical step) from a radiotherapy linear accelerator. The resulting Cherenkov light emission – and thus luminescent probe excitation – occurred exclusively along the irradiation plane due to a short diffusion path of secondary particles and Cherenkov photons. Cherenkov-excited luminescence was detected orthogonally to the sheet beam by gated, intensified charge coupled device camera. Analogously to light sheet microscopy, a series of luminescence images was taken for varied axial positions (depths) of the Cherenkov light sheet in sample. Knowledge of the excitation plane position allowed a 3D image stack deconvolution and depth-variant attenuation correction. The 3D image post-processing yielded a true spatial density distribution of luminescent molecules in highly scattering tissue. Results: We recovered a three-dimensional shape and position of 400 µL lesion-mimicking phantom tubes containing 25 µM solution of PtG4 molecular probe from 3 centimeter deep tissue-like media. The high sensitivity of CELSI also allowed resolving 100 micron capillaries of test solution. Functional information of partial oxygen pressure at the site of PtG4 molecular probe was recovered from luminescence lifetime CELSI. Finally, in-vivo sheet beam CELSI localized milimeter-sized PtG4-labelled tumor phantoms in multiple biological objects (hairless mice) from single scan. Conclusion: Presented sheet beam CELSI technique greatly extended the useful depth range of luminescence molecular imaging. More importantly, the light sheet microscopy approach was successfully adapted to CELSI, providing means to recover a completely attenuation-corrected 3D image of luminescent probe distribution. Gated CELSI acquisition yielded functional information of a spatially resolved oxygen concentration map of deep lying targets. This work was supported by NIH research grant R01CA109558 and R21EB017559, as well as by Pilot Grant Funds from the Norris Cotton Cancer Center.

60 APPLIED LIFE SCIENCES↗

Mathematical nuances of Gaussian process-driven autonomous experimentation

Abstract The fields of machine learning (ML) and artificial intelligence (AI) have transformed almost every aspect of science and engineering. The excitement for AI/ML methods is in large part due to their perceived novelty, as compared to traditional methods of statistics, computation, and applied mathematics. But clearly, all methods in ML have their foundations in mathematical theories, such as function approximation, uncertainty quantification, and function optimization. Autonomous experimentation is no exception; it is often formulated as a chain of off-the-shelf tools, organized in a closed loop, without emphasis on the intricacies of each algorithm involved. The uncomfortable truth is that the success of any ML endeavor, and this includes autonomous experimentation, strongly depends on the sophistication of the underlying mathematical methods and software that have to allow for enough flexibility to consider functions that are in agreement with particular physical theories. We have observed that standard off-the-shelf tools, used by many in the applied ML community, often hide the underlying complexities and therefore perform poorly. In this paper, we want to give a perspective on the intricate connections between mathematics and ML, with a focus on Gaussian process-driven autonomous experimentation. Although the Gaussian process is a powerful mathematical concept, it has to be implemented and customized correctly for optimal performance. We present several simple toy problems to explore these nuances and highlight the importance of mathematical and statistical rigor in autonomous experimentation and ML. One key takeaway is that ML is not, as many had hoped, a set of agnostic plug-and-play solvers for everyday scientific problems, but instead needs expertise and mastery to be applied successfully. Graphical abstract

97 MATHEMATICS AND COMPUTING↗