In this work we propose and analyze a new framework to learn feedback control policies that exhibit provable guarantees on the closed-loop performance and robustness to bounded (adversarial) perturbations. These policies are learned from expert demonstrations without any prior knowledge of the task, its cost function, and system dynamics. In contrast to the existing algorithms in imitation learning and inverse reinforcement learning, we use a Lipschitz-constrained loss minimization scheme to learn control policies with certified robustness. We establish robust stability of the closed-loop system under the learned control policy and derive an upper bound on its regret, which bounds the sub-optimality of the closed-loop performance with respect to the expert policy. We also derive a robustness bound for the deterioration of the closed-loop performance under bounded (adversarial) perturbations on the state measurements. Ultimately, our results suggest the existence of an underlying tradeoff between nominal closed-loop performance and adversarial robustness, and that improvements in nominal closed-loop performance can only be made at the expense of robustness to adversarial perturbations. Numerical results validate our analysis and demonstrate the effectiveness of our robust feedback policy learning framework.

The advent of larger machine learning (ML) models have improved state-of-the-art (SOTA) performance in various modeling tasks, ranging from computer vision to natural language. As ML models continue increasing in size, so does their respective energy consumption and computational requirements. However, the methods for tracking, reporting, and comparing energy consumption remain limited. We presentEnergyVis, an interactive energy consumption tracker for ML models. Consisting of multiple coordinated views, EnergyVis enables researchers to interactively track, visualize and compare model energy consumption across key energy consumption and carbon footprint metrics (kWh and CO2), helping users explore alternative deployment locations and hardware that may reduce carbon footprints. EnergyVis aims to raise awareness concerning computational sustainability by interactively highlighting excessive energy usage during model training; and by providing alternative training options to reduce energy usage.

In the literature, many methods have been proposed The extended state observer (ESO) is an inherent component for tuning bandwidth-parameterized observers. In [49] of the robust control framework that relies on the cancellation and [38], the authors presented an analytical tuning method providing of disturbances using their lumped estimate in the feedforward the best performance of the ADRC structure expressed component of the robust control law. The general idea of solely upon the control-error-dependent criteria in a noiseless such control structure was utilized in many specific robust algorithms environment. In [39] and [6], the authors considered also the such as active disturbance rejection control (ADRC) control cost as a factor that needs to be minimized to reduce the [13, 54], disturbance observer based control (DOBC) [5, 23], energy consumption of the robust control process, while in [26] or robust observer based control [16], while its applicability the observation error of the measured signals was taken into has been proven in many fields including power electronics account. Tuning procedures described in [28] and [12] have [20, 25, 46], temperature control [53], motion control [32, 42], utilized prior knowledge about the plant structure and some and robotics [30, 22]. Besides the fact that there is a wide variety known or identified model parameters to obtain assumed control of ESO architectures that deal with disadvantages of a most performance requirements. In [28] and [14], the authors commonly used Luenberger-like extended high-gain observer presented an observer tuning method that is relative to gains (HGO) [52, 8] in terms of the general disturbance observation of the selected ADRC controller. Some methods consider automatic quality [34], transient performance [40], or the robustness to tools designed for tuning the overall ADRC structure, measurement noise [37, 21], the final characteristics of the control including observer gains, to satisfy some predefined criteria determining system performance depend greatly on the appropriate tuning the robustness of the control structure [36].

Structural health monitoring (SHM) systems use the non-destructive testing principle for damage identification. As part of SHM, the propagation of ultrasonic guided waves (UGWs) is tracked and analyzed for the changes in the associated wave pattern. These changes help identify the location of a structural damage, if any. We advance existing research by accounting for uncertainty in the material and geometric properties of a structure. The physics model used in this study comprises of a monolithically coupled system of acoustic and elastic wave equations, known as the wave propagation in fluid-solid and their interface (WpFSI) problem. As the UGWs propagate in the solid, fluid, and their interface, the wave signal displacement measurements are contrasted against the benchmark pattern. For the numerical solution, we develop an efficient algorithm that successfully addresses the inherent complexity of solving the multiphysics problem under uncertainty. We present a procedure that uses Gaussian process regression and convolutional neural network for predicting the UGW propagation in a solid-fluid and their interface under uncertainty. First, a set of training images for different realizations of the uncertain parameters of the inclusion inside the structure is generated using a monolithically-coupled system of acoustic and elastic wave equations. Next, Gaussian processes trained with these images are used for predicting the propagated wave with convolutional neural networks for further enhancement to produce high-quality images of the wave patterns for new realizations of the uncertainty. The results indicate that the proposed approach provides an accurate prediction for the WpFSI problem in the presence of uncertainty.

Tomographic imaging refers to the reconstruction of a 3D object from its 2D projections by sectioning the object, through the use of any kind of penetrating wave, from many different directions. It has had a revolutionary impact in a number of fields ranging from biology, physics, and chemistry to astronomy [1, 2]. The technique requires an accurate image reconstruction, however, and the resulting reconstruction problem is an ill-posed optimization problem because of insufficient measurements [3]. A direct consequence of ill-posedness is that the reconstruction does not have a unique solution. Therefore, quantifying the solution quality is challenging, given the absence of ground truth and the presence of measurement noise. Moreover, ill-posedness creates a requirement for regularization that imports new parameters to the problem. Regularization parameter choice can lead to substantial variations in reconstruction, and ascertaining optimal values of such parameters is difficult without availing oneself of ground truth [4]. The transition from an optimization perspective on tomographic inversion to a Bayesian statistical perspective can provide a useful reframing of these issues. In particular, the ill-posedness of the optimization view can be replaced by quantified uncertainty in the statistical view, whereas regularization now appears in the guise of parameter estimation.

The potential of digital twin technology is immense, specifically in the infrastructure, aerospace, and automotive sector. However, practical implementation of this technology is not at an expected speed, specifically because of lack of application-specific details. In this paper, we propose a novel digital twin framework for stochastic nonlinear multi-degree of freedom (MDOF) dynamical systems. The approach proposed in this paper strategically decouples the problem into two time-scales -- (a) a fast time-scale governing the system dynamics and (b) a slow time-scale governing the degradation in the system. The proposed digital twin has four components - (a) a physics-based nominal model (low-fidelity), (b) a Bayesian filtering algorithm a (c) a supervised machine learning algorithm and (d) a high-fidelity model for predicting future responses. The physics-based nominal model combined with Bayesian filtering is used combined parameter state estimation and the supervised machine learning algorithm is used for learning the temporal evolution of the parameters. While the proposed framework can be used with any choice of Bayesian filtering and machine learning algorithm, we propose to use unscented Kalman filter and Gaussian process. Performance of the proposed approach is illustrated using two examples. Results obtained indicate the applicability and excellent performance of the proposed digital twin framework.

The National Energy Research Scientific Computing Center at Lawrence Berkeley National Laboratory, in collaboration with the Argonne Leadership Computing Facility, is partnering with UK-based Codeplay Software to enhance GPU compiler capabilities for NVIDIA. This collaboration will help NERSC and ALCF users, along with researchers in the high-performance computing community, to produce high-performance applications that are portable across compute architectures from multiple vendors. Today, most artificial intelligence software, including for cars, is developed using graphics processors designed for video games, according to Codeplay. The company provides tools designed to enable software to be accelerated by graphics processors or the latest specialized AI processors. NERSC supercomputers are used for scientific research by researchers working in diverse areas such as alternative energy, environment, high-energy and nuclear physics, advanced computing, materials science and chemistry.

Machine learning is a field of computational science which first emerged in the 1950s. However, our ability to effectively harness the power of machine learning techniques was only truly realised in the 1990s. In ecology, the earliest adoption of machine learning came about in the early 2000s, when regression tree algorithms were applied to spatial data to predict species distributions. This was quickly adapted in the field of marine ecology to study the distribution of many pelagic species. Since that time, machine learning algorithms have been adapted and applied in various studies in the marine environment, from population models, image recognition, and experimental studies.

Pre-training models such as BERT have achieved great success in many natural language processing tasks. However, how to obtain better sentence representation through these pre-training models is still worthy to exploit. Previous work has shown that the anisotropy problem is an critical bottleneck for BERT-based sentence representation which hinders the model to fully utilize the underlying semantic features. Therefore, some attempts of boosting the isotropy of sentence distribution, such as flow-based model, have been applied to sentence representations and achieved some improvement. In this paper, we find that the whitening operation in traditional machine learning can similarly enhance the isotropy of sentence representations and achieve competitive results. Furthermore, the whitening technique is also capable of reducing the dimensionality of the sentence representation. Our experimental results show that it can not only achieve promising performance but also significantly reduce the storage cost and accelerate the model retrieval speed.

To help researchers better analyze the massive amount of data they collect from their experiments, the Department of Energy is dedicating $29 million to develop new machine learning tools and advanced algorithms that will benefit multiple scientific fields and inform cutting-edge solutions for a variety of complex problems. Today's scientific facilities, instruments and high-performance computing (HPC) simulations regularly generate terabytes of data -- so much that traditional analysis methods can struggle to interpret the data efficiently. More advanced machine learning tools can identify patterns in data that humans cannot detect, running up to thousands of times faster than traditional data analysis techniques. "As research tools like computers or microscopes have gotten more powerful, the amount of data they can gather has gotten overwhelming--and scientists need new capabilities to make sense of it all," Energy Secretary Jennifer M. Granholm said. "Advanced analysis methods will help them unlock the full potential behind all this data, so that we can solve even our most complex challenges."