Search NASA⌕ Search

SEARCH · Search NASA

Results for “adaptive learning”

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 649 records · Page 36

A transfer learning approach to energy-efficient control of small and medium-sized commercial buildings

Model-free reinforcement learning (RL) provides a data-driven and adaptive approach to optimize building energy use while satisfying occupant comfort. This powerful tool does not need any prior knowledge about the environment and system it is optimizing and can adapt its policy based on the changes in captures. Like any other data-driven tool, it faces high training costs due to the extensive agent-environment interactions required to capture long-term building dynamics and user comfort. Transfer learning, particularly policy distillation, offers a promising way to accelerate training by leveraging pretrained RL agents in different building and system types. Here, this study investigates online student distillation, in which the student model updates its neural network weights using outputs from teacher models. The work introduces a student distillation strategy designed for efficient knowledge transfer, along with a teacher selection method that ensures high-quality guidance. The approach is validated using a highly calibrated whole building energy model for a small/medium commercial building test facility. Results show substantial reductions in training time and data requirements while surpassing the performance of ASHRAE Guideline 36, an advanced rule-based control strategy. The distilled RL model required 45% less data and achieved 20% higher cumulative rewards than a state-of-the-art RL model, with faster convergence and lower energy consumption. These outcomes demonstrate that effective transfer learning enables a scalable and data-efficient energy management solution for commercial buildings.

ASHRAE guideline 36↗

SIDDA: SInkhorn Dynamic Domain Adaptation for image classification with equivariant neural networks

Modern neural networks (NNs) often do not generalize well in the presence of a ‘covariate shift’; that is, in situations where the training and test data distributions differ, but the conditional distribution of classification labels given the data remains unchanged. In such cases, NN generalization can be reduced to a problem of learning more robust, domain-invariant features. Domain adaptation (DA) methods include a broad range of techniques aimed at achieving this; however, these methods have struggled with the need for extensive hyperparameter tuning, which then incurs significant computational costs. In this work, we introduce SInkhorn Dynamic Domain Adaptation (SIDDA), an out-of-the-box DA training algorithm built upon the Sinkhorn divergence, that can achieve effective domain alignment with minimal hyperparameter tuning and computational overhead. We demonstrate the efficacy of our method on multiple simulated and real datasets of varying complexity, including simple shapes, handwritten digits, real astronomical observations, and remote sensing data. These datasets exhibit covariate shifts due to noise, blurring, differences between telescopes, and variations in imaging wavelengths. SIDDA is compatible with a variety of NN architectures, and it works particularly well in improving classification accuracy and model calibration when paired with symmetry-aware equivariant NNs (ENNs). We find that SIDDA consistently enhances the generalization capabilities of NNs, achieving up to a ${\approx}40\%$ improvement in classification accuracy on unlabeled target data, while also providing a more modest performance gain of $\lesssim 1\%$ on labeled source data. We also study the efficacy of DA on ENNs with respect to the varying group orders of the dihedral group DN, and find that the model performance improves as the degree of equivariance increases. Finally, if SIDDA achieves proper domain alignment, it also enhances model calibration on both source and target data, with the most significant gains in the unlabeled target domain—achieving over an order of magnitude improvement in the expected calibration error and Brier score. SIDDA’s versatility across various NN models and datasets, combined with its automated approach to domain alignment, has the potential to significantly advance multi-dataset studies by enabling the development of highly generalizable models.

79 ASTRONOMY AND ASTROPHYSICS↗

Protein Conformational States—A First Principles Bayesian Method

Automated identification of protein conformational states from simulation of an ensemble of structures is a hard problem because it requires teaching a computer to recognize shapes. We adapt the naïve Bayes classifier from the machine learning community for use on atom-to-atom pairwise contacts. The result is an unsupervised learning algorithm that samples a ‘distribution’ over potential classification schemes. We apply the classifier to a series of test structures and one real protein, showing that it identifies the conformational transition with >95% accuracy in most cases. A nontrivial feature of our adaptation is a new connection to information entropy that allows us to vary the level of structural detail without spoiling the categorization. This is confirmed by comparing results as the number of atoms and time-samples are varied over 1.5 orders of magnitude. Further, the method’s derivation from Bayesian analysis on the set of inter-atomic contacts makes it easy to understand and extend to more complex cases.

97 MATHEMATICS AND COMPUTING↗

Optoelectronic analogs of self-programming neural nets - Architecture and methodologies for implementing fast stochastic learning by simulated annealing

Self-organization and learning is a distinctive feature of neural nets and processors that sets them apart from conventional approaches to signal processing. It leads to self-programmability which alleviates the problem of programming complexity in artificial neural nets. In this paper architectures for partitioning an optoelectronic analog of a neural net into distinct layers with prescribed interconnectivity pattern to enable stochastic learning by simulated annealing in the context of a Boltzmann machine are presented. Stochastic learning is of interest because of its relevance to the role of noise in biological neural nets. Practical considerations and methodologies for appreciably accelerating stochastic learning in such a multilayered net are described. These include the use of parallel optical computing of the global energy of the net, the use of fast nonvolatile programmable spatial light modulators to realize fast plasticity, optical generation of random number arrays, and an adaptive noisy thresholding scheme that also makes stochastic learning more biologically plausible. The findings reported predict optoelectronic chips that can be used in the realization of optical learning machines.

Farhat, Nabil H.↗

A Hybrid Reinforcement Learning and Cellular Automata Model for Crowd Simulation on the GPU

We present a GPU-based hybrid model for crowd simulations. The model uses reinforcement learning to guide groups of pedestrians towards a goal while adapting to environmental dynamics, and a cellular automaton to describe individual pedestrians’ interactions. In contrast to traditional multi-agent reinforcement learning methods, our model encodes the learned navigation policy into a navigation map, which is used by the cellular automaton’s update rule to calculate the next simulation step. As a result, reinforcement learning is independent of the number of agents, allowing the simulation of large crowds. Implementation of this model on the GPU allows interactive simulations of several hundreds of pedestrians.

Ruiz, Sergio↗

FedCSpc: A Cross-Silo Federated Learning System With Error-Bounded Lossy Parameter Compression

Cross-Silo federated learning is widely used for scaling deep neural network (DNN) training over data silos from different locations worldwide while guaranteeing data privacy. Communication has been identified as the main bottleneck when training large-scale models due to large-volume model parameters and gradient transmission across public networks with limited bandwidth. Most previous works focus on gradient compression, while limited work tries to compress parameters that can not be ignored and extremely affect communication performance during the training. Here, to bridge this gap, we propose FedCSpc: an efficient cross-silo federated learning system with an XAI-driven adaptive parameter compression strategy for large-scale model training. Our work substantially differs from existing gradient compression techniques due to the distinct data features of gradient and parameter. The key contributions of this paper are fourfold. (1) Our designed FedCSpc proposes to compress the parameter during the training using the state-of-the-art error-bounded lossy compressor – SZ3. (2) We develop an adaptive compression error bound adjustment algorithm to guarantee the model accuracy effectively. (3) We exploit an efficient approach to utilize the idle CPU resources of clients to compress the parameters. (4) We perform a comprehensive evaluation with a wide range of models and benchmarks on a GPU cluster with 65 GPUs. Results show that FedCSpc can achieve the same model accuracy as FedAvg while reducing the data volume of parameters and gradients in communication by up to 7.39× and 288×, respectively. With 32 clients on a 4 Gb size model, FedCSpc significantly outperforms FedAvg in wall-clock time in the emulated WAN environment (at the bandwidth of 1 Gbps or lower without loss of generality).

SZ3↗

A User-Centered Approach to Adaptive Hypertext Based on an Information Relevance Model

Rapid and effective to information in large electronic documentation systems can be facilitated if information relevant in an individual user's content can be automatically supplied to this user. However most of this knowledge on contextual relevance is not found within the contents of documents, it is rather established incrementally by users during information access. We propose a new model for interactively learning contextual relevance during information retrieval, and incrementally adapting retrieved information to individual user profiles. The model, called a relevance network, records the relevance of references based on user feedback for specific queries and user profiles. It also generalizes such knowledge to later derive relevant references for similar queries and profiles. The relevance network lets users filter information by context of relevance. Compared to other approaches, it does not require any prior knowledge nor training. More importantly, our approach to adaptivity is user-centered. It facilitates acceptance and understanding by users by giving them shared control over the adaptation without disturbing their primary task. Users easily control when to adapt and when to use the adapted system. Lastly, the model is independent of the particular application used to access information, and supports sharing of adaptations among users.

Mathe, Nathalie↗

Learning linear optical circuits with coherent states

We analyze the energy and training data requirements for supervised learning of an M-mode linear optical circuit by minimizing an empirical risk defined solely from the action of the circuit on coherent states. When the linear optical circuit acts non-trivially only on k < M unknown modes (i.e. a linear optical k-junta), we provide an energy-efficient, adaptive algorithm that identifies the junta set and learns the circuit. We compare two schemes for allocating a total energy, E, to the learning algorithm. In the first scheme, each of the T random training coherent states has energy E/T. In the second scheme, a single random MT-mode coherent state with energy E is partitioned into T training coherent states. The latter scheme exhibits a polynomial advantage in training data size sufficient for convergence of the empirical risk to the full risk due to concentration of measure on the $(2MT-1)$-sphere. Specifically, generalization bounds for both schemes are proven, which indicate that for ε-approximation of the full risk by the empirical risk with high probability, $O(E^{2/3}M^{2/3}/\epsilon^{2/3})$ training states are sufficient for the first scheme and $O(E^{1/3}M^{1/3}/\epsilon^{2/3})$ training states are sufficient for the second scheme.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Bayesian sparse learning with preconditioned stochastic gradient MCMC and its applications

Deep neural networks have been successfully employed in an extensive variety of research areas, including solving partial differential equations. Despite its significant success, there are some challenges in effectively training DNN, such as avoiding overfitting in over-parameterized DNNs and accelerating the optimization in DNNs with pathological curvature. Here, we propose a Bayesian type sparse deep learning algorithm. The algorithm utilizes a set of spike-and-slab priors for the parameters in the deep neural network. The hierarchical Bayesian mixture will be trained using an adaptive empirical method. That is, one will alternatively sample from the posterior using preconditioned stochastic gradient Langevin Dynamics (PSGLD), and optimize the latent variables via stochastic approximation. The sparsity of the network is achieved while optimizing the hyperparameters with adaptive searching and penalizing. A popular SG-MCMC approach is Stochastic gradient Langevin dynamics (SGLD). However, considering the complex geometry in the model parameter space in nonconvex learning, updating parameters using a universal step size in each component as in SGLD may cause slow mixing. To address this issue, we apply a computationally manageable preconditioner in the updating rule, which provides a step-size parameter to adapt to local geometric properties. Moreover, by smoothly optimizing the hyperparameter in the preconditioning matrix, our proposed algorithm ensures a decreasing bias, which is introduced by ignoring the correction term in the preconditioned SGLD. According to the existing theoretical framework, we show that the proposed algorithm can asymptotically converge to the correct distribution with a controllable bias under mild conditions. Numerical tests are performed on both synthetic regression problems and learning solutions of elliptic PDE, which demonstrate the accuracy and efficiency of the present work.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Deep Learning Emulation of Atmospheric Correction for Geostationary Sensors

New generation geostationary satellites make reflectance observations available at a continental scale with unprecedented spatiotemporal resolution and spectral range. Generating Earth monitoring products from these observations requires retrieval of the basic parameter, surface reflectance (SR), by atmospheric correction (AC). Algorithms for atmospheric correction, including Multi-Angle Implementation of Atmospheric Correction (MAIAC), are adapted for each sensor and are too computationally complex to be run in real time, relying instead on look-up tables with precomputed values. Machine learning methods, including convolutional neural networks, have demonstrated performance in learning complex, nonlinear mappings and extracting insight from high-dimensional remote sensing data. In this work, we present a deep learning emulator of MAIAC to retrieve both SR and cloud products. Using this adaptation of deep learning-based emulation to remote sensing, we demonstrate stable SR retrieval over a variety of land covers and viewing conditions and accurate cloud detection. Further, a comparison of computation time suggests emulation as a compelling alternative for expensive physical simulation, especially for applications benefited by near-real time data, such as agricultural management and disaster response.

Duffy, Kate↗

An Edge-Cloud Integrated Solution for Buildings Demand Response Using Reinforcement Learning

Buildings, as major energy consumers, can provide great untapped demand response (DR) resources for grid services. However, their participation remains low in real-life. One major impediment for popularizing DR in buildings is the lack of cost-effective automation systems that can be widely adopted. Existing optimization-based smart building control algorithms suffer from high costs on both building-specific modeling and on-demand computing resources. To tackle these issues, this paper proposes a cost-effective edge-cloud integrated solution using reinforcement learning (RL). Beside RL’s ability to solve sequential optimal decision-making problems, its adaptability to easy-to-obtain building models and the off-line learning feature are likely to reduce the controller’s implementation cost. Using a surrogate building model learned automatically from building operation data, an RL agent learns an optimal control policy on cloud infrastructure, and the policy is then distributed to edge devices for execution. Simulation results demonstrate the control efficacy and the learning efficiency in buildings of different sizes. A preliminary cost analysis on a 4-zone commercial building shows the annual cost for optimal policy training is only 2.25% of the DR incentive received. Results of this study show a possible approach with higher return on investment for buildings to participate in DR programs.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Using Board Games as Subject Matter for Developing Expertise in Model-Based Systems Engineering

As more organizations transition from traditional document-centric systems engineering to a model-based approach, many are challenged to train their staff in new languages, tools, and methodologies, while managing the expectations of stakeholders and their expected model outcomes. In particular, challenges associated with learning a new modeling language and developing skills in the 'art' of modeling present organizations with formidable obstacles to realizing this transition. This paper hypothesizes that systems engineers may more readily learn how to correctly model with SysML, and develop intuition about the art of modeling and using patterns, if their learning references a commonly and thoroughly-understood subject, such as a board game. This paper presents a case for the use of board games as subject matter for new modelers. It demonstrates the concept with a sample model of Hasbro's popular board game, Monopoly, and discusses the limitations of this approach and potential adaptations that may broaden the applicability of the learned skills to projects. Finally, results from a small feasibility assessment and concepts for more formal study to evaluate the hypothesis are presented.

Model-Based Systems Engineering↗

Using Board Games as Subject Matter for Developing Expertise in Model-Based Systems Engineering

As more organizations transition from traditional document-centric systems engineering to a model-based approach, many are challenged to train their staff in new languages, tools, and methodologies, and manage the expectations of stakeholders and their expected model outcomes. In particular, challenges associated with learning a new modeling language and developing skills in the 'art' of modeling present organizations with formidable obstacles to realizing this transition. This paper hypothesizes that systems engineers may more readily learn how to correctly model with SysML, and develop intuition about the art of modeling and using patterns, if their learning references a commonly and thoroughly-understood subject matter, such as a board game. This paper presents a case for the use of board games as subject matter for new modelers, demonstrates the concept with a sample model of Hasbro's popular board game, Monopoly, and discusses the limitations of this approach and potential adaptations that may broaden the applicability of the learned skills to projects.

Systems Engineering↗

Active Learning of Microgrid Frequency Dynamics Using Neural Ordinary Differential Equations

Accurate frequency modelling of inverter‐based resource (IBR)‐dominated power systems is crucial for ensuring stable, reliable and resilient operations, particularly given their inherent low‐inertia characteristics and fast dynamics that traditional swing equation‐based models inadequately capture. This paper explores neural ordinary differential equations (Neural ODEs) as a computationally efficient, data‐driven framework for modelling power system frequency dynamics, specifically within microgrids integrating high penetrations of distributed energy resources (DERs). The developed neural ODEs framework incorporates a neural network architecture designed to capture input dynamics. By actively perturbing the system with a known signal, the Python‐based neural ODEs framework was trained using measured system states and inputs, without the need for detailed system information. The framework, tested on a model of the Cordova, AK, microgrid, achieved a goodness of fit ranging from 60% to 99% across different state variables and maintained a mean square error in the 10 -6 p.u. range under square and step excitation signals. The proposed approach demonstrated robustness to measurement noise and initial condition variations while maintaining low computational complexity suitable for real‐time power system control applications. Furthermore, transfer learning enabled the neural ODEs model to adapt to the following changes in system topology or generator dispatch, highlighting its effectiveness for dynamic microgrids with frequently evolving configurations and diverse DERs.

Aryal, Tara [South Dakota State Univ., Brookings, ↗

Adaptive enhanced sampling by force-biasing using neural networks

A machine learning assisted method is presented for molecular simulation of systems with rugged free energy landscapes. The method is general and can be combined with other advanced sampling techniques. In the particular implementation proposed here, it is illustrated in the context of an adaptive biasing force approach where, rather than relying on discrete force estimates, one can resort to a self-regularizing artificial neural network to generate continuous, estimated generalized forces. By doing so, the proposed approach addresses several shortcomings common to adaptive biasing force and other algorithms. Specifically, the neural network enables smooth estimates of generalized forces in sparsely sampled regions, force estimates in previously unexplored regions, and continuous force estimates with which to bias the simulation, as opposed to biases generated at specific points of a discrete grid. The usefulness of the method is illustrated with three different examples, chosen to highlight the wide range of applicability of the underlying concepts. In all three cases, the new method is found to enhance considerably the underlying traditional adaptive biasing force approach. Furthermore, the method is also found to provide improvements over previous implementations of neural network assisted algorithms.

97 MATHEMATICS AND COMPUTING↗

Overview of the KSTAR experiments toward fusion reactor

The Korean Superconducting Tokamak Advanced Research has been focused on exploring the key physics and engineering issues for future fusion reactors by demonstrating the long pulse operation of high beta steady-state discharge. Advanced scenarios are being developed with the goal for steady-state operation, and significant progress has been made in high ℓ i , hybrid and high beta scenarios with β N of 3. In the new operation scenario called fast ion regulated enhanced (FIRE), fast ions play an essential role in confinement enhancement. GK simulations show a significant reduction of the thermal energy flux when the thermal ion fraction decreases and the main ion density gradient is reversed by the fast ions in FIRE mode. Optimization of 3D magnetic field techniques, including adaptive control and real-time machine learning control algorithm, enabled long-pulse operation and high-performance ELM-suppressed discharge. Symmetric multiple shattered pellet injections (SPIs) and real-time disruption event characterization and forecasting are being performed to mitigate and avoid the disruptions associated with high-performance, long-pulse ITER-like scenarios. Finally, the near-term research plan will be addressed with the actively cooled tungsten divertor, a major upgrade of the NBI and helicon current drive heating, and transition to a full metallic wall.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Further adoption of conservation tillage can increase maize yields in the western US Corn Belt

Conservation tillage can reduce soil erosion, increase soil health, and decrease labor and fuel input costs. Despite these benefits, potential yield impacts remain an important concern for farmers considering adoption. Previous research suggests that conservation tillage is likely to have the largest yield benefits in more arid conditions, but a lack of field-level analyses across climatic, management and soil conditions limits confidence in such predictions. Satellite imagery provides the opportunity to monitor agricultural lands at sub-field resolution across large spatial scales and wide environmental gradients. Here we investigate the maize yield impacts of conservation tillage in the semi-arid western US Corn Belt, using sub-field resolution datasets on tillage practices and crop yields derived from satellite data spanning four states (Nebraska, Kansas, South Dakota, and North Dakota) between 2008 and 2020. On these datasets, we estimate heterogenous yield outcomes for several thousand maize fields across gradients in climate, soil quality and irrigation status by using a causal forests analysis, an adaptation of the random forests machine-learning algorithm for causal inference on observational data. We find that long-term adoption of conservation tillage increased rainfed maize yields by an average of 9.9% in the region. Impacts on irrigated yields were small and not statistically significant. These results, along with an analysis of variables related to greater than average yield benefits, indicate that improved water infiltration and retention are the primary reasons for conservation tillage benefits. Despite yield benefits, many fields estimated to see increased yields under long term low till have not adopted the practice. Therefore, we identify specific counties likely to benefit most from increased levels of adoption. Our results strengthen the understanding of the impacts of conservation agriculture on crop yields and help define environments and counties most likely to benefit from conservation tillage.

54 ENVIRONMENTAL SCIENCES↗

Decision-tree structures utilizing a phase-transition material

The rich internal physics due to competing electronic phases present in phase-transition materials such as VO2 offer the potential for compact building block design for emerging non-von Neumann computing technologies. Here, based on the relaxation dynamics of an insulator-metal phase transition, we demonstrate experimentally a decision-tree classifier embedded within a single volatile resistive switching device. The tree is constructed by the combination of the voltage pulse and relaxation time and can adapt to different tasks. We use machine learning to analyze the relaxation process, enabling a predictive voltage-relaxation time phase diagram for the electrical resistance state. Classification of the etiology of the chronic cough is presented as a proof-of-principle use case. Further, our approach can be generalized to broader classes of solid-state and solid-liquid interfacial systems that demonstrate a variety of phase relaxations.

36 MATERIALS SCIENCE↗