Search NASA⌕ Search

SEARCH · Search NASA

Results for “stochastic gradient”

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 73 records · Page 4

Finding MIDDLE Ground: Scalable and Secure Distributed Learning

Edge computing methods allow devices to efficiently train a high-performing, robust, and personalized model for predictive tasks. However, these methods succumb to privacy and scalability concerns such as adversarial data recovery and expensive model communication. Furthermore, edge computing methods unrealistically assume that all devices train an identical model. In practice, edge devices have varying computational and memory constraints which may not allow certain devices to have the space or speed to train a specific model. To overcome these issues, we propose MIDDLE: a model independent distributed learning algorithm which allows heterogeneous edge devices to assist each other’s training while communicating only non-sensitive information. MIDDLE unlocks the ability for edge devices, regardless of computational or memory constraints, to assist each other even with completely different model architectures. Furthermore, MIDDLE does not require model or gradient communication which greatly reduces communication size and time. We prove that MIDDLE attains the optimal convergence rate O(1/sqrt(TM)) of stochastic gradient descent for convex and non-convex smooth optimization (for total iterations T and batch size M). Finally, our experimental results demonstrate that MIDDLE (even in non-IID data settings) attains robust and high-performing models without model or gradient communication.

Bornstein, Marc I.↗

Learning generative neural networks with physics knowledge

Deep generative neural networks have enabled modeling complex distributions, but incorporating physics knowledge into the neural networks is still challenging and is at the core of current physics-based machine learning research. To this end, we propose a physics generative neural network (PhysGNN), a new class of generative neural networks for learning unknown distributions in a physical system described by partial differential equations (PDE). PhysGNN couples PDE systems with generative neural networks. It is a fully differentiable model that allows back-propagation of gradients through both numerical PDE solvers and generative neural networks, and is trained by minimizing the discrete Wasserstein distance between generated and observed probability distributions of the PDE outputs using the stochastic gradient descent method. Moreover, PhysGNN does not require adversarial training like standard generative neural networks, which offers better stability than adversarial training. We show that PhysGNN can learn complex distributions in stochastic inverse problems, where conventional methods such as maximum likelihood estimation and momentum matching methods may be inapplicable when little knowledge is known about the form of unknown distributions or the physical model is too complex. Furthermore, our method allows physics-based generative neural network training for learning complex distributions in the context of differential equations.

97 MATHEMATICS AND COMPUTING↗

Efficient Generalizable Deep Learning

Stochastic optimization is a fundamental field of research for machine learning. Stochastic gradient descent (SGD) and related methods provide a feasible means to train complicated prediction models over large datasets. SGD, however, does not explicitly address the problem of overfitting, which can lead to predictions that perform poorly on new data. This difference between loss performance on unseen testing data verses that of training data defines the generalization gap of a model. We introduce a new computational kernel called Stochastic Hessian Projection (SHP) that uses a maximum likelihood framework to simultaneously estimate gradient noise covariance and local curvature of the loss function. Our analysis illustrates that these quantities affect the evolution of parameter uncertainty and therefore generalizability. We show how these computations allow us to predict the generalization gap without requiring holdout data. Explicitly assessing this metric for generalizability during training may improve machine learning predictions when data is scarce and understanding prediction variability is critical.

42 ENGINEERING↗

Semiautomatic registration of 3D transabdominal ultrasound images for patient repositioning during postprostatectomy radiotherapy

Purpose: The aim of the present work is to propose and evaluate registration algorithms of three-dimensional (3D) transabdominal (TA) ultrasound (US) images to setup postprostatectomy patients during radiation therapy. Methods: Three registration methods have been developed and evaluated to register a reference 3D-TA-US image acquired during the planning CT session and a 3D-TA-US image acquired before each treatment session. The first method (method A) uses only gray value information, whereas the second one (method B) uses only gradient information. The third one (method C) combines both sets of information. All methods restrict the comparison to a region of interest computed from the dilated reference positioning volume drawn on the reference image and use mutual information as a similarity measure. The considered geometric transformations are translations and have been optimized by using the adaptive stochastic gradient descent algorithm. Validation has been carried out using manual registration by three operators of the same set of image pairs as the algorithms. Sixty-two treatment US images of seven patients irradiated after a prostatectomy have been registered to their corresponding reference US image. The reference registration has been defined as the average of the manual registration values. Registration error has been calculated by subtracting the reference registration from the algorithm result. For each session, the method has been considered a failure if the registration error was above both the interoperator variability of the session and a global threshold of 3.0 mm. Results: All proposed registration algorithms have no systematic bias. Method B leads to the best results with mean errors of −0.6, 0.7, and −0.2 mm in left–right (LR), superior–inferior (SI), and anterior–posterior (AP) directions, respectively. With this method, the standard deviations of the mean error are of 1.7, 2.4, and 2.6 mm in LR, SI, and AP directions, respectively. The latter are inferior to the interoperator registration variabilities which are of 2.5, 2.5, and 3.5 mm in LR, SI, and AP directions, respectively. Failures occur in 5%, 18%, and 10% of cases in LR, SI, and AP directions, respectively. 69% of the sessions have no failure. Conclusions: Results of the best proposed registration algorithm of 3D-TA-US images for postprostatectomy treatment have no bias and are in the same variability range as manual registration. As the algorithm requires a short computation time, it could be used in clinical practice provided that a visual review is performed.

60 APPLIED LIFE SCIENCES↗

End-to-End Differentiable Modeling and Management of the Environment

Focal Area: (2) Data acquisition and assimilation enabled by machine learning, AI, and advanced methods including experimental/network design/optimization. We emphasize the importance of leveraging optimization techniques from AI/machine learning (ML) to solve challenging problems in Earth system modeling. Science Challenge: Automatic differentiation has had a transformative effect on ML by allowing the calculation of gradients of arbitrary functions in an incredibly large class of models. We can potentially realize similar improvements in parameter estimation and control for Earth system models (ESMs) by reimplementing them in computational frameworks from ML. Practitioners working with large (>10 7 parameters) models in ML and AI can obtain good predictive performance in a range of spatiotemporal tasks by making use of optimization via stochastic gradient descent and incorporating prior knowledge at multiple levels. We propose writing ESMs in open-source computational frameworks such as Torch, Tensorflow, and JAX to greatly expand the scope of environmental forecasting and management challenges, which can be addressed by leveraging automatic differentiation and gradient descent-like algorithms. We do not call for a wholesale replacement of physical models with data-driven surrogates, but rather advocate for interleaving physical and empirical equations in a manner that is most faithful to the extent of our scientific knowledge and observational data. Central to this topic is the merging of differentiable physical simulations with differentiable optimization layers, which are now both beginning to come to the forefront.

54 ENVIRONMENTAL SCIENCES↗

Improving Deep Neural Networks’ Training for Image Classification With Nonlinear Conjugate Gradient-Style Adaptive Momentum

Momentum is crucial in stochastic gradient-based optimization algorithms for accelerating or improving training deep neural networks (DNNs). In deep learning practice, the momentum is usually weighted by a well-calibrated constant. However, tuning the hyperparameter for momentum can be a significant computational burden. In this article, we propose a novel adaptive momentum for improving DNNs training; this adaptive momentum, with no momentum-related hyperparame- ter required, is motivated by the nonlinear conjugate gradient (NCG) method. Stochastic gradient descent (SGD) with this new adaptive momentum eliminates the need for the momentum hyperparameter calibration, allows using a significantly larger learning rate, accelerates DNN training, and improves the final accuracy and robustness of the trained DNNs. For example, SGD with this adaptive momentum reduces classification errors for training ResNet110 for CIFAR10 and CIFAR100 from 5.25% to 4.64% and 23.75% to 20.03%, respectively. Furthermore, SGD, with the new adaptive momentum, also benefits adversarial training and, hence, improves the adversarial robustness of the trained DNNs.

97 MATHEMATICS AND COMPUTING↗

WE-AB-209-09: Optimization of Rotational Arc Station Parameter Optimized Radiation Therapy

Purpose: To develop a fast optimization method for station parameter optimized radiation therapy (SPORT) and show that SPORT is capable of improving VMAT in both plan quality and delivery efficiency. Methods: The angular space from 0° to 360° was divided into 180 station points (SPs). A candidate aperture was assigned to each of the SPs based on the calculation results using a column generation algorithm. The weights of the apertures were then obtained by optimizing the objective function using a state-of-the-art GPU based Proximal Operator Graph Solver (POGS) within seconds. Apertures with zero or low weight were thrown out. To avoid being trapped in a local minimum, a stochastic gradient descent method was employed which also greatly increased the convergence rate of the objective function. The above procedure repeated until the plan could not be improved any further. A weighting factor associated with the total plan MU also indirectly controlled the complexities of aperture shapes. The number of apertures for VMAT and SPORT was confined to 180. The SPORT allowed the coexistence of multiple apertures in a single SP. The optimization technique was assessed by using three clinical cases (prostate, H&N and brain). Results: Marked dosimetric quality improvement was demonstrated in the SPORT plans for all three studied cases. Prostate case: the volume of the 50% prescription dose was decreased by 22% for the rectum. H&N case: SPORT improved the mean dose for the left and right parotids by 15% each. Brain case: the doses to the eyes, chiasm and inner ears were all improved. SPORT shortened the treatment time by ∼1 min for the prostate case, ∼0.5 min for brain case, and ∼0.2 min for the H&N case. Conclusion: The superior dosimetric quality and delivery efficiency presented here indicates that SPORT is an intriguing alternative treatment modality.

60 APPLIED LIFE SCIENCES↗

In silico prediction of toxicity of non-congeneric industrial chemicals using ensemble learning based modeling approaches

Ensemble learning approach based decision treeboost (DTB) and decision tree forest (DTF) models are introduced in order to establish quantitative structure–toxicity relationship (QSTR) for the prediction of toxicity of 1450 diverse chemicals. Eight non-quantum mechanical molecular descriptors were derived. Structural diversity of the chemicals was evaluated using Tanimoto similarity index. Stochastic gradient boosting and bagging algorithms supplemented DTB and DTF models were constructed for classification and function optimization problems using the toxicity end-point in T. pyriformis. Special attention was drawn to prediction ability and robustness of the models, investigated both in external and 10-fold cross validation processes. In complete data, optimal DTB and DTF models rendered accuracies of 98.90%, 98.83% in two-category and 98.14%, 98.14% in four-category toxicity classifications. Both the models further yielded classification accuracies of 100% in external toxicity data of T. pyriformis. The constructed regression models (DTB and DTF) using five descriptors yielded correlation coefficients (R{sup 2}) of 0.945, 0.944 between the measured and predicted toxicities with mean squared errors (MSEs) of 0.059, and 0.064 in complete T. pyriformis data. The T. pyriformis regression models (DTB and DTF) applied to the external toxicity data sets yielded R{sup 2} and MSE values of 0.637, 0.655; 0.534, 0.507 (marine bacteria) and 0.741, 0.691; 0.155, 0.173 (algae). The results suggest for wide applicability of the inter-species models in predicting toxicity of new chemicals for regulatory purposes. These approaches provide useful strategy and robust tools in the screening of ecotoxicological risk or environmental hazard potential of chemicals. - Graphical abstract: Importance of input variables in DTB and DTF classification models for (a) two-category, and (b) four-category toxicity intervals in T. pyriformis data. Generalization and predictive abilities of the constructed (c) DTB and (d) DTF regression models to predict the T. pyriformis toxicity of diverse chemicals. - Highlights: • Ensemble learning (EL) based models constructed for toxicity prediction of chemicals • Predictive models used a few simple non-quantum mechanical molecular descriptors. • EL-based DTB/DTF models successfully discriminated toxic and non-toxic chemicals. • DTB/DTF regression models precisely predicted toxicity of chemicals in multi-species. • Proposed EL based models can be used as tool to predict toxicity of new chemicals.

60 APPLIED LIFE SCIENCES↗

Prediction of gas hydrate saturation using machine learning and optimal set of well-logs

We report resistivity and acoustic logs are widely used to estimate gas hydrate saturation in various sedimentary systems using one of the two popular methods ((1) acoustic velocity and (2) electrical resistivity), but the limitations of these two methods are often overlooked, which include (i) well-specific calibration of empirical exponents in the electrical resistivity method, (ii) assumption of known pore morphology for gas hydrates in the acoustic velocity method, and (iii) presence of unknown mineralogy and bulk modulus terms in the acoustic velocity method. NMR-density porosity-derived gas hydrate saturation based on the analysis of the transverse magnetization relaxation time (T2) is considered the most precise method, but acquisition of NMR-based logs is limited at relatively recent drilled sites; additionally, its use in conventional oil and gas reservoirs is not that common due to higher cost and operational deployment limitations associated with acquiring NMR well-logs. This study proposes a new method that predicts gas hydrate saturation (S h ) for any well using porosity, bulk density, and compressional wave (P wave) velocity well-logs with neural network (or stochastic gradient descent regression) without any well-specific calibration and/or other aforementioned shortcomings of the existing methods. The method is developed by examining the underlying dependency between S h and different combinations of well-logs, chosen from 6 routine logs, with 12 different machine learning (ML) algorithms. The accuracy of the proposed method in predicting S h is ~ 84%, which is better than the accuracy of seismic and electrical resistivity methods (≤ 75%) per the results reported by three different studies. The robustness of the method in the specific case of permafrost-associated gas hydrates is demonstrated with well-log data from two wells drilled on the Alaska North Slope.

58 GEOSCIENCES↗

A phase transition for finding needles in nonlinear haystacks with LASSO artificial neural networks

To fit sparse linear associations, a LASSO sparsity inducing penalty with a single hyperparameter provably allows to recover the important features (needles) with high probability in certain regimes even if the sample size is smaller than the dimension of the input vector (haystack). More recently learners known as artificial neural networks (ANN) have shown great successes in many machine learning tasks, in particular fitting nonlinear associations. Small learning rate, stochastic gradient descent algorithm and large training set help to cope with the explosion in the number of parameters present in deep neural networks. Yet few ANN learners have been developed and studied to find needles in nonlinear haystacks. Driven by a single hyperparameter, our ANN learner, like for sparse linear associations, exhibits a phase transition in the probability of retrieving the needles, which we do not observe with other ANN learners. To select our penalty parameter, we generalize the universal threshold of Donoho and Johnstone (Biometrika 81(3):425–455, 1994) which is a better rule than the conservative (too many false detections) and expensive cross-validation. In the spirit of simulated annealing, we propose a warm-start sparsity inducing algorithm to solve the high-dimensional, non-convex and non-differentiable optimization problem. We perform simulated and real data Monte Carlo experiments to quantify the effectiveness of our approach.

97 MATHEMATICS AND COMPUTING↗

Growing grasses in unprofitable areas of US Midwest croplands could increase species richness

The US has large potential to grow perennial energy crops, but because these crops are rarely grown in current agricultural landscapes, it is unclear how biodiversity may be affected. Over time, as agriculture has increased, many grassland species have declined. In addition, not all agricultural land is profitable for growing annual crops. Unprofitable areas were responsible for a loss of approximately $110 million USD per year from 2013 to 2016. Based on this, we want to know how converting less-profitable portions of agricultural fields to switchgrass, a native prairie grass, would influence species occurrence. To address this question, we developed an alternative landscape in which clustered corn/soy acres with a low return on investment (ROI) were replaced with grassland. We also developed and validated species distribution models to predict changes in species occurrence for 28 avian species in Iowa in response to landscape management. Furthermore, we compared results for three different models: Random forest (RF), Stochastic gradient boosting (GBM), and Neural network (Nnet) and found that all models performed well and predicted similar species distribution. Predicted species richness increased by 3.66% (RF), 2.79% (GBM), and 7.51% (Nnet) when we simulated a change in management for ~3% of Iowa's low ROI corn/soybean areas to grassland. If harvested, these areas could generate approximately 7.6 million dry tons/year of switchgrass for bioenergy, thereby increasing farmers earnings. Unprofitable areas tended to occur along streams, which suggest that incorporating partially harvested riparian buffers can benefit avian biodiversity, while improving water quality and reducing unnecessary costs for farmers.

59 BASIC BIOLOGICAL SCIENCES↗

Accelerating Hamiltonian Monte Carlo for Bayesian inference in neural networks and neural operators

Hamiltonian Monte Carlo (HMC) is a powerful and accurate method to sample from the posterior distribution in Bayesian inference. However, HMC techniques are computationally demanding for Bayesian neural networks due to the high dimensionality of the network’s parameter space and the non-convexity of their posterior distributions. Therefore, various approximation techniques, such as variational inference (VI) or stochastic gradient MCMC, are often employed to infer the posterior distribution of the network parameters. Such approximations introduce inaccuracies in the inferred distributions, resulting in unreliable uncertainty estimates. In this work, we propose a hybrid approach that combines inexpensive VI and accurate HMC methods to efficiently and accurately quantify uncertainties in neural networks and neural operators. The proposed approach leverages an initial VI training on the full network. We examine the influence of individual parameters on the prediction uncertainty, which shows that a large proportion of the parameters do not contribute substantially to uncertainty in the network predictions. This information is then used to significantly reduce the dimension of the parameter space, and HMC is performed only for the subset of network parameters that strongly influence prediction uncertainties. This yields a framework for accelerating the full batch HMC for posterior inference in neural networks. We demonstrate the efficiency and accuracy of the proposed framework on deep neural networks and operator networks, showing that inference can be performed for large networks with tens to hundreds of thousands of parameters. Finally, we show that this method can effectively learn surrogates for complex physical systems by modeling the operator that maps from upstream conditions to wall-pressure data on a cone in hypersonic flow.

Bayesian inference↗

DeepONet-grid-UQ: A trustworthy deep operator framework for predicting the power grid’s post-fault trajectories

This paper proposes a novel data-driven method for the reliable prediction of the power grid’s post-fault trajectories, i.e., the power grid’s dynamic response after a disturbance or fault. Here, the proposed method is based on the recently proposed concept of Deep Operator Networks (DeepONets). Unlike traditional neural networks that learn to approximate functions, DeepONets are designed to approximate nonlinear operators, i.e., mappings between infinite-dimensional spaces. Under this operator framework, we design a novel and efficient DeepONet that (i) takes as inputs the trajectories collected before and during the fault and (ii) outputs the predicted post-fault trajectories. In addition, we endow our method with the much-needed ability to balance efficiency with reliable/trustworthy predictions via uncertainty quantification. To this end, we propose and compare two novel methods that enable quantifying the predictive uncertainty. First, we propose a Bayesian DeepONet (B-DeepONet) that uses stochastic gradient Hamiltonian Monte-Carlo to sample from the posterior distribution of the DeepONet trainable parameters. Then, we design a Probabilistic DeepONet (Prob-DeepONet) that uses a probabilistic training strategy to enable quantifying uncertainty at virtually no extra computational cost. Finally, we validate the proposed methods’ predictive power and uncertainty quantification capability using the New York-New England power grid model.

24 POWER TRANSMISSION AND DISTRIBUTION↗

HPC formulations of optimization algorithms for tensor completion

Tensor completion is a powerful tool used to estimate or recover missing values in multi-way data. It has seen great success in domains such as product recommendation and healthcare. Tensor completion is most often accomplished via low-rank sparse tensor factorization, a computationally expensive non-convex optimization problem which has only recently been studied in the context of parallel computing. In this work, we study three optimization algorithms that have been successfully applied to tensor completion: alternating least squares (ALS), stochastic gradient descent (SGD), and coordinate descent (CCD++). We explore opportunities for parallelism on shared- and distributed-memory systems and address challenges such as memory- and operation-efficiency, load balance, cache locality, and communication. Among our advancements are a communication-efficient CCD++ algorithm, an ALS algorithm rich in level-3 BLAS routines, and an SGD algorithm which combines stratification with asynchronous communication. Furthermore, we show that introducing randomization during ALS and CCD++ can accelerate convergence. We evaluate our parallel formulations on a variety of real datasets on a modern supercomputer and demonstrate speedups through 16384 cores. Further, these improvements reduce time-to-solution from hours to seconds on real-world datasets. We show that after our optimizations, ALS is advantageous on parallel systems of small-to-moderate scale, while both ALS and CCD++ provide the lowest time-to-solution on large-scale distributed systems.

97 MATHEMATICS AND COMPUTING↗

Extracting Grain Orientations from EBSD Patterns of Polycrystalline Materials Using Convolutional Neural Networks

Abstract We present a deep learning approach to the indexing of electron backscatter diffraction (EBSD) patterns. We design and implement a deep convolutional neural network architecture to predict crystal orientation from the EBSD patterns. We design a differentiable approximation to the disorientation function between the predicted crystal orientation and the ground truth; the deep learning model optimizes for the mean disorientation error between the predicted crystal orientation and the ground truth using stochastic gradient descent. The deep learning model is trained using 374,852 EBSD patterns of polycrystalline nickel from simulation and evaluated using 1,000 experimental EBSD patterns of polycrystalline nickel. The deep learning model results in a mean disorientation error of 0.548° compared to 0.652° using dictionary based indexing.

Materials Science↗

Concurrent multi-peak Bragg coherent x-ray diffraction imaging of 3D nanocrystal lattice displacement via global optimization

Abstract In this paper we demonstrated a method to reconstruct vector-valued lattice distortion fields within nanoscale crystals by optimization of a forward model of multi-reflection Bragg coherent diffraction imaging (MR-BCDI) data. The method flexibly accounts for geometric factors that arise when making BCDI measurements, is amenable to efficient inversion with modern optimization toolkits, and allows for globally constraining a single image reconstruction to multiple Bragg peak measurements. This is enabled by a forward model that emulates the multiple Bragg peaks of a MR-BCDI experiment from a single estimate of the 3D crystal sample. We present this forward model, we implement it within the stochastic gradient descent optimization framework, and we demonstrate it with simulated and experimental data of nanocrystals with inhomogeneous internal lattice displacement. We find that utilizing a global optimization approach to MR-BCDI affords a reliable path to convergence of data which is otherwise challenging to reconstruct.

36 MATERIALS SCIENCE↗

Stable diagonal stripes in the t-J model at $\bar{n}$ h = 1/8 doping from fPEPS calculations

We investigate the two-dimensional t–J model at a hole doping of $\bar{n}$ h= 1 / 8 using recently developed high accuracy fermionic projected entangled pair states method. By applying the stochastic gradient descent method combined with the Monte Carlo sampling technique, we obtain the ground state hole energy Ehole = -1.621 for J/t = 0.4. We show that the ground state has stable diagonal stripes instead of vertical stripes with a width of 4 unit cells, and stripe filling ρ l = 0.5. We further show that the long-range superconductivity order is suppressed at this point.

36 MATERIALS SCIENCE↗

Integrative analysis of transcriptomic and proteomic data of Shewanella oneidensis: missing value imputation using temporal datasets

Despite significant improvements in recent years, proteomic datasets currently available still suffer large number of missing values. Integrative analyses based upon incomplete proteomic and transcriptomic da-tasets could seriously bias the biological interpretation. In this study, we applied a non-linear data-driven stochastic gradient boosted trees (GBT) model to impute missing proteomic values for proteins experi-mentally undetected, using a temporal transcriptomic and proteomic dataset of Shewanella oneidensis. In this dataset, genes expression was measured after the cells were exposed to 1 mM potassium chromate for 5-, 30-, 60-, and 90-min, while protein abundance was measured only for 45- and 90-min samples. With the goal of elucidating the relationship between temporal gene expression and protein abundance data, and then using it to impute missing proteomic values for samples of 45-min (which does not have cognate transcriptomic data) and 90-min, we initially used nonlinear Smoothing Splines Curve Fitting (SSCF) to identify temporal relationships among transcriptomic data at different time points and then imputed missing gene expression measurements for the sample at 45-min. After the imputation was validated by biological constrains (i.e. operons), we used a data-driven Gradient Boosted Trees (GBT) model to uncover possible non-linear relationships between temporal transcriptomic and proteomic data, and to impute protein abundance for the proteins experimentally undetected in the 45- and 90-min sam-ples, based on relevant predictors such as temporal mRNA gene expression data, cellular roles, molecular weight, sequence length, protein length, guanine-cytosine (GC) content and triple codon counts. The imputed protein values were validated using biological constraints such as operon, regulon and pathway information. Finally, we demonstrated that such missing value imputation improved characterization of the temporal response of S. oneidensis to chromate.

SBR mercury↗