Search NASA⌕ Search

SEARCH · Search NASA

Results for “network inference algorithms”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

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

At least 127 records · Page 7
All sourcesNASA NTRSDOE OSTI
DOE OSTI2020

nPINNs: nonlocal Physics-Informed Neural Networks for a parametrized nonlocal universal Laplacian operator. Algorithms and Applications

Physics-informed neural networks (PINNs) are effective in solving inverse problems based on differential and integro-differential equations with sparse, noisy, unstructured, and multifidelity data. PINNs incorporate all available information, including governing equations (reflecting physical laws), initial-boundary conditions, and observations of quantities of interest, into a loss function to be minimized, thus recasting the original problem into an optimization problem. In this paper, we extend PINNs to parameter and function inference for integral equations such as nonlocal Poisson and nonlocal turbulence models, and we refer to them as nonlocal PINNs (nPINNs). The contribution of the paper is three-fold. First, we propose a unified nonlocal Laplace operator, which converges to the classical Laplacian as one of the operator parameters, the nonlocal interaction radius δ goes to zero, and to the fractional Laplacian as δ goes to infinity. This universal operator forms a super-set of classical Laplacian and fractional Laplacian operators and, thus, has the potential to fit a broad spectrum of data sets. We provide theoretical convergence rates with respect to δ and verify them via numerical experiments. Second, we use nPINNs to estimate the two parameters, δ and α, characterizing the kernel of the unified operator. The strong non-convexity of the loss function yielding multiple (good) local minima reveals the occurrence of the operator mimicking phenomenon, that is, different pairs of estimated parameters could produce multiple solutions of comparable accuracy. Third, we propose another nonlocal operator with spatially variable order α(γ), which is more suitable for modeling turbulent Couette flow. Our results show that nPINNs can jointly infer this function as well as δ. More importantly, these parameters exhibit a universal behavior with respect to the Reynolds number, a finding that contributes to our understanding of nonlocal interactions in wall-bounded turbulence.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2020

Physics for neuromorphic computing

Neuromorphic computing takes inspiration from the brain to create energy-efficient hardware for information processing, capable of highly sophisticated tasks. Systems built with standard electronics achieve gains in speed and energy by mimicking the distributed topology of the brain. Scaling-up such systems and improving their energy usage, speed and performance by several orders of magnitude requires a revolution in hardware. Here, we discuss how including more physics in the algorithms and nanoscale materials used for data processing could have a major impact in the field of neuromorphic computing. We review striking results that leverage physics to enhance the computing capabilities of artificial neural networks, using resistive switching materials, photonics, spintronics and other technologies. We discuss the paths that could lead these approaches to maturity, towards low-power, miniaturized chips that could infer and learn in real time.

device physics↗
DOE OSTI2024

Ultrafast jet classification at the HL-LHC

Abstract Three machine learning models are used to perform jet origin classification. These models are optimized for deployment on a field-programmable gate array device. In this context, we demonstrate how latency and resource consumption scale with the input size and choice of algorithm. Moreover, the models proposed here are designed to work on the type of data and under the foreseen conditions at the CERN large hadron collider during its high-luminosity phase. Through quantization-aware training and efficient synthetization for a specific field programmable gate array, we show that O ( 100 ) ns inference of complex architectures such as Deep Sets and Interaction Networks is feasible at a relatively low computational resource cost.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
DOE OSTI2020

Sensing Service Transformer Secondary Currents using Planar Magnetic Pick-up Coils

The rapid deployment of distributed energy resources (DERs) is creating stresses on utility assets in the distribution networks. The service transformer, which is the most commonly used utility asset is often not monitored, since the cost of available sensing solutions is as high as the transformer itself. This paper presents a method to use an array of printed circuit board coils to intelligently monitor North American pole top service transformers. The approach, uses a novel, self-calibrating algorithm to infer the fixed geometry of the installation, by recording data over an extended period of time. The proposed method can enable rapid installation of a non-intrusive sensor that can be used to monitor loading levels of the distribution transformer. Furthermore, these sensors can help in planning, prognostics and advanced situational awareness for utilities.

24 POWER TRANSMISSION AND DISTRIBUTION↗
DOE OSTI2023

Risk-Informed Condition Evaluation of Solar-centered Energy Generation and Distribution Networks through Bayesian Learning and Inference

We develop a methodology based on Bayesian inference over Probabilistic Graphical Models (PGMs) to understand and quantify risk in solar-centered grids using targeted measurements and learned system behavior. Being non-prescriptive but, rather, able to infer system behavior and, ultimately, address risk queries from data, our machine learning-type paradigm is tailored for diverse topologies and threat scenarios often associated with distributed energy generation and photovoltaic distributed energy resources (PV-DERs) in particular. We describe algorithmic processes for: (i) learning the structure of PGMs that result from attack-prone PV-DER-proliferated distribution systems, (ii) quantifying cause-effect relationships, and (iii) evaluating risk queries based on diverse evidence. The contributions are illustrated on a residential grid subject to output impairment attacks on its PV-DER infrastructure.

24 POWER TRANSMISSION AND DISTRIBUTION↗
DOE OSTI2023

A Hardware and Software Co-design Framework for Energy Efficient Neuromorphic Systems

Neuromorphic systems can be realized by a variety of algorithms and architectures. A common understanding is that spiking neuromorphic designs, which encode information into spatio-temporal spiking events, are both a biologically-accurate and efficient way of processing information. However, representing the information through timing relationships induces sophisticated circuit designs in traditional CMOS-based implementations. In recent years, high-capacity resistive memory (RRAM, aka, memristor) has demonstrated great potential in mimicking synaptic behaviors. Several RRAM-based spiking neuromorphic designs exist, most of which focus on rate coding schemes. These designs simplify circuit implementations of neuron models and explore challenges such as unsatisfactory speed, resolution, and performance. As an alternative, we will explore temporal coding spiking neuromorphic systems that encode information as the relative timing of neuron activations (spikes), which have been proven to be more adaptive and energy-efficient. Developing a neuromorphic system for spiking neural network (SNN) inference and online training, however, faces some major technical challenges: (1) It lacks circuit implementation support for temporal-coding SNN to achieve satisfying power efficiency and accuracy; (2) Although existing research works have investigated memristive synapse and neuron designs for spike-timing-dependent plasticity, the non-ideal conditions in implementation, such as device variations and signal degradation, degrade online learning accuracy of large scale systems; and (3) Non-optimized, inter-layer data traffic in SNNs, leads to unnecessary data communication costs. In this project, we plan to address these challenges by a hardware and software co-design framework that incorporates solutions at the circuit, architecture, and algorithm levels. At the circuit-level, we will elaborate on the in-situ SNN processing element designs for supporting both inference and online training modes. Variation-aware schemes will be studied to improve reliability. At the architecture level, we propose a pipelined, asynchronous architecture to retain the timing resolution of spikes. At the algorithm level, we will investigate an innovative SNN training algorithm for enabling activation sparsification and reducing unnecessary data communication costs. This neuromorphic system will provide an effective solution to real-life energy-constrained applications and significantly contribute to the exploration of next-generation high-performance computing systems under the DOE context.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2021

Particle image velocimetry analysis with simultaneous uncertainty quantification using Bayesian neural networks

Particle image velocimetry (PIV) is an effective tool in experimental fluid mechanics for extracting flow fields from images. Recently, convolutional neural networks (CNNs) have been used to perform PIV analysis with accuracy on par with classical methods. Here we extend the use of CNNs to analyze PIV data while providing simultaneous uncertainty quantification on the inferred flow field. The method we apply in this paper is a Bayesian convolutional neural network (BCNN) which learns distributions of the CNN weights through variational Bayes. In order to demonstrate the utility of BCNNs for the PIV task, we compare the performance of three distinct BCNN models with simple architectures. The first network estimates flow velocity from image interrogation regions only. Our second model learns to infer velocity from both the image interrogation regions and interrogation region cross-correlation maps. Finally, our best performing network infers velocities from interrogation region cross-correlation maps only. We find that BCNNs using interrogation region cross-correlation maps as inputs perform better than those using interrogation windows only as inputs and discuss reasons why this may be the case. Additionally, we test the best performing BCNN on a full synthetic test image pair and a real image pair from the 1st International PIV Challenge. We show that ~98% of true particle displacements from the full synthetic image pair can be captured within the BCNN's 95% confidence intervals, and that the BCNN's performance on the real image pair is quantitatively similar to that of algorithms tested in the 1st International PIV Challenge. Finally, we show that BCNNs can be generalized to be used with multi-pass PIV algorithms with a moderate loss in accuracy, which may be overcome by future work on finetuning and training schemes. So to our knowledge, this is the first use of Bayesian neural networks to perform PIV.

47 OTHER INSTRUMENTATION↗
NASA NTRS2000

Qualitative Discovery in Medical Databases

Implication rules have been used in uncertainty reasoning systems to confirm and draw hypotheses or conclusions. However a major bottleneck in developing such systems lies in the elicitation of these rules. This paper empirically examines the performance of evidential inferencing with implication networks generated using a rule induction tool called KAT. KAT utilizes an algorithm for the statistical analysis of empirical case data, and hence reduces the knowledge engineering efforts and biases in subjective implication certainty assignment. The paper describes several experiments in which real-world diagnostic problems were investigated; namely, medical diagnostics. In particular, it attempts to show that: (1) with a limited number of case samples, KAT is capable of inducing implication networks useful for making evidential inferences based on partial observations, and (2) observation driven by a network entropy optimization mechanism is effective in reducing the uncertainty of predicted events.

Maluf, David A.↗
DOE OSTI2025

A resolution independent neural operator

The Deep operator network (DeepONet) is a powerful yet simple neural operator architecture that utilizes two deep neural networks to learn mappings between infinite-dimensional function spaces. This architecture is highly flexible, allowing the evaluation of the solution field at any location within the desired domain. However, it imposes a strict constraint on the input space, requiring all input functions to be discretized at the same locations; this limits its practical applications. Here, in this work, we introduce a general framework for operator learning from input–output data with arbitrary number and locations of sensors. This begins by introducing a resolution-independent DeepONet (RI-DeepONet), enabling it to handle input functions that are arbitrarily, but sufficiently finely, discretized. To this end, we propose two dictionary learning algorithms to adaptively learn a set of appropriate continuous basis functions, parameterized as implicit neural representations (INRs), from correlated signals defined on arbitrary point cloud data. These basis functions are then used to project arbitrary input function data as a point cloud onto an embedding space (i.e., a vector space of finite dimensions) with dimensionality equal to the dictionary size, which can be directly used by DeepONet without any architectural changes. In particular, we utilize sinusoidal representation networks (SIRENs) as trainable INR basis functions. The introduced dictionary learning algorithms are then used in a similar way to learn an appropriate dictionary of basis functions for the output function data, which defines a new neural operator architecture referred to as the R esolution I ndependent N eural O perator (RINO). In the RINO, the operator learning task simplifies to learning a mapping from the coefficients of input basis functions to the coefficients of output basis functions. We demonstrate the robustness and applicability of RINO in handling arbitrarily (but sufficiently richly) sampled input and output functions during both training and inference through several numerical examples.

Deep operator network (DeepONet)↗
DOE OSTI2024

Sentinel

Network intrusion detection systems (NIDS) are commonplace in network security but they frequently employ algorithms that are computational demanding requiring hardware and software with significant power requirements. Two examples of such resource-intensive algorithms used for network security are regular expression matching and broader signature pattern matching which are commonly used in deep packet inspection (DPI). Network security algorithms that have large power requirements may be a challenge for low-power internet-of-things (IoT) environments, which generally lack the power resources to implement complex security measures like computationally expensive DPI at the edge. Furthermore, IoT environments incorporating 5G standalone networks have network latency constraints beyond just power that make DPI at the edge even more difficult. Programmable logic is ideally suited for machine learning inference for DPI because of its deep instruction level parallelism and single-cycle memory access. Machine learning approaches for DPI have been explored before using the programmable logic of field programmable gate arrays (FPGA) as a potential solution for NIDS approaches that would be power-suitable for IoT. However, those previous programmable logic NIDS approaches utilize either a supervised or unsupervised learning model. Sentinel utilizes the ensemble of these two machine learning approaches known as a semi-supervised approach which has shown promise in NIDS implementations. Sentinel provides a programmable logic implementation of a semi-supervised approach for DPI which operates at much lower power and latency than a GPU implementation with negligible loss of accuracy due to quantization through a logistic regressor.

Anderson, MatthewW [Idaho National Laboratory (INL↗
DOE OSTI2026

Real-Time Detection of Charge Jumps in Superconducting Qubits with a Convolutional Neural Network

Ionizing radiation from cosmic rays and gammas can induce discontinuous jumps in the environmental charge of superconducting qubits (charge jumps), causing correlated errors that challenge fault-tolerant quantum computing while simultaneously providing a detection signature for quantum sensing applications. Current detection methods operate offline, introducing latency incompatible with in-the-loop qubit control. In this paper, an online detector of charge jumps for superconducting qubits, based on a dilated causal convolutional neural network (DCCNN) designed for in-the-loop deployment on the Quantum Instrumentation Control Kit (QICK) platform, is presented. The network is trained on synthetic Ramsey tomography scans generated from qubit templates measured at the Northwestern Experimental Underground Site (NEXUS) at Fermilab, and translated to FPGA firmware via hls4ml with ap_fixed$\langle 16,6 \rangle$ quantization, reaching a per-inference latency of $6.19 μ$s on the Zynq UltraScale+ RFSoC ZCU216. At this operating point the DCCNN matches the detection efficiency of the established offline $χ^2$ algorithm ($0.843 \pm 0.022$ vs. $0.866 \pm 0.020$ on $|Δq| \in [0.1, 0.5] e$ at matched false-positive rate), while requiring no per-qubit hyperparameter tuning. This shifts charge-jump detection from a post-hoc diagnostic to a control-loop primitive, enabling adaptive protocols that respond to radiation-induced events in situ, with applications to quantum-computing error mitigation and to the use of superconducting qubits as particle detectors.

Gaytan-Villarreal, Daniel [Carnegie Mellon U.]↗
NASA NTRS2016

A Neural Network Approach to Infer Optical Depth of Thick Ice Clouds at Night

One of the roadblocks to continuously monitoring cloud properties is the tendency of clouds to become optically black at cloud optical depths (COD) of 6 or less. This constraint dramatically reduces the quantitative information content at night. A recent study found that because of their diffuse nature, ice clouds remain optically gray, to some extent, up to COD of 100 at certain wavelengths. Taking advantage of this weak dependency and the availability of COD retrievals from CloudSat, an artificial neural network algorithm was developed to estimate COD values up to 70 from common satellite imager infrared channels. The method was trained using matched 2007 CloudSat and Aqua MODIS data and is tested using similar data from 2008. The results show a significant improvement over the use of default values at night with high correlation. This paper summarizes the results and suggests paths for future improvement.

Minnis, P.↗
DOE OSTI2024

EFIT-Prime: Probabilistic and physics-constrained reduced-order neural network model for equilibrium reconstruction in DIII-D

We introduce EFIT-Prime, a novel machine learning surrogate model for EFIT (Equilibrium FIT) that integrates probabilistic and physics-informed methodologies to overcome typical limitations associated with deterministic and ad hoc neural network architectures. EFIT-Prime utilizes a neural architecture search-based deep ensemble for robust uncertainty quantification, providing scalable and efficient neural architectures that comprehensively quantify both data and model uncertainties. Physically informed by the Grad–Shafranov equation, EFIT-Prime applies a constraint on the current density J tor and a smoothness constraint on the first derivative of the poloidal flux, ensuring physically plausible solutions. Furthermore, the spatial location of the diagnostics is explicitly incorporated in the inputs to account for their spatial correlation. Extensive evaluations demonstrate EFIT-Prime's accuracy and robustness across diverse scenarios, most notably showing good generalization on negative-triangularity discharges that were excluded from training. Timing studies indicate an ensemble inference time of 15 ms for predicting a new equilibrium, offering the possibility of plasma control in real-time, if the model is optimized for speed.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2023

Advanced data analysis in inertial confinement fusion and high energy density physics

Bayesian analysis enables flexible and rigorous definition of statistical model assumptions with well-characterized propagation of uncertainties and resulting inferences for single-shot, repeated, or even cross-platform data. This approach has a strong history of application to a variety of problems in physical sciences ranging from inference of particle mass from multi-source high-energy particle data to analysis of black-hole characteristics from gravitational wave observations. The recent adoption of Bayesian statistics for analysis and design of high-energy density physics (HEDP) and inertial confinement fusion (ICF) experiments has provided invaluable gains in expert understanding and experiment performance. In this Review, we discuss the basic theory and practical application of the Bayesian statistics framework. We highlight a variety of studies from the HEDP and ICF literature, demonstrating the power of this technique. Due to the computational complexity of multi-physics models needed to analyze HEDP and ICF experiments, Bayesian inference is often not computationally tractable. Two sections are devoted to a review of statistical approximations, efficient inference algorithms, and data-driven methods, such as deep-learning and dimensionality reduction, which play a significant role in enabling use of the Bayesian framework. We provide additional discussion of various applications of Bayesian and machine learning methods that appear to be sparse in the HEDP and ICF literature constituting possible next steps for the community. We conclude by highlighting community needs, the resolution of which will improve trust in data-driven methods that have proven critical for accelerating the design and discovery cycle in many application areas.

47 OTHER INSTRUMENTATION↗
DOE OSTI2023

Deep-learning-aided extraction of optical constants in scanning near-field optical microscopy

Scanning near-field optical microscopy is one of the most effective techniques for spectroscopy of nanoscale systems. However, inferring optical constants from the measured near-field signal can be challenging because of a complicated and highly nonlinear interaction between the scanned probe and the sample. Conventional fitting methods applied to this problem often suffer from the lack of convergence or require human intervention. Here, we develop an alternative approach where the optical parameter extraction is automated by a deep learning network. The network provides an initial estimate that is subsequently refined by a traditional fitting algorithm. In conclusion, we show that this method demonstrates superior accuracy, stability against noise, and computational speed when applied to simulated near-field spectra.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗
NASA NTRS1995

Optical fiber sensors for damage analysis in aerospace materials

Under this grant, fiber optic sensors were investigated for use in the nondestructive evaluation of aging aircraft. Specifically, optical fiber sensors for detection and location of impacts on a surface, and for detection of corrosion in metals were developed. The use of neural networks was investigated for determining impact location by processing the output of a network of fiberoptic strain sensors distributed on a surface. This approach employs triangulation to determine location by comparing the arrival times at several sensors, of the acoustic signal generated by the impact. For this study, a neural network simulator running on a personal computer was used to train a network using a back-propagation algorithm. Fiber optic extrinsic Fabry-Perot interferometer (EFPI) strain sensors are attached to or embedded in the surface, so that stress waves emanating from an impact can be detected. The ability of the network to determine impact location by time-or-arrival of acoustic signals was assessed by comparing network outputs with actual experimental results using impacts on a panel instrumented with optical fiber sensors. Using the neural network to process the sensor outputs, the impact location can be inferred to centimeter range accuracy directly from the arrival time data. In addition, the network can be trained to determine impact location, regardless of material anisotropy. Results demonstrate that a back-propagation network identifies impact location for an anisotropic graphite/bismaleimide plate with the same accuracy as that for an isotropic aluminum plate. Two different approaches were investigated for the development of fiber optic sensors for corrosion detection in metals, both utilizing optical fiber sensors with metal coatings. In the first approach, an extrinsic Fabry-Perot interferometric fiber optic strain sensor was placed under tensile stress, and while in the resulting strained position, a thick coating of metal was applied. Due to an increase in the quantity of material, the sensor does not return to its original position upon removal of the applied stress, and some residual strain is maintained within the sensor element. As the metal thickness decreases due to corrosion, this strain is released, providing the sensing mechanism for corrosion detection. In the second approach, photosensitive optical fibers with long period Bragg gratings in the core were coated with metal. The Bragg gratings serve to couple core modes at discrete wavelengths to cladding modes. Since cladding modes interact with the metal coating surrounding the fiber cladding, the specific wavelengths coupled from core to cladding depend on the refractive index of the metal coating. Therefore, as the metal corrodes, the resulting change in index of the coating may be measured by measuring the change in wavelength of the coupled mode. Results demonstrate that both approaches can be successfully used to track the loss in metal coating on the optical fiber sensors due to corrosion.

Schindler, Paul↗
DOE OSTI2026

Identification of drug repurposing candidates for amyotrophic lateral sclerosis using electronic health records: a retrospective cohort study

Amyotrophic lateral sclerosis (ALS) is a progressive neurodegenerative disease with a life expectancy of only 3–5 years and few approved treatments. To identify drug repurposing candidates for the treatment of ALS, we analysed the electronic health records (EHRs) of a large cohort of military veterans with ALS. We analysed the EHRs of individuals in the US Veterans Health Administration (VHA) database who were diagnosed with ALS between Jan 1, 2009 and Dec 31, 2019 to assess medication effects. Individuals without recorded prescriptions after the date of diagnosis were excluded. Two sets of criteria were applied to ascertain exposure. Exposure criteria A were met if the dispense date or the end date of the medication was within 12 months of ALS diagnosis and the end date was at least 6 months after the dispense date. Exposure criteria B were met if there were at least two dispenses within 6 months before diagnosis and 12 months after diagnosis. Propensity score-matched control groups were generated on the basis of confounders included in the EHR, with methodology of potential outcomes used to infer treatment effects. The primary outcome was death. A standard Cox proportional hazards analysis was done to assess association with survival. Survival was defined as the time from diagnosis date recorded in the EHR to death reported in the Department for Veterans Affairs Vital Status File. Follow-up survival time was censored on Dec 31, 2020, for those alive on this date. Downstream protein targets of drugs with clinically significant effects were analysed using the protein–protein interaction networks-based algorithm PathFX. The EHRs of 11 003 individuals with ALS in the VHA database were appropriate for analysis. 162 medications with treatment groups of 30 or more individuals were identified. Among these 162 medications, 27 were associated with statistically significant changes (≥0·1) in the hazard ratio (HR) for death. 18 of the medications were associated with a reduced HR for death (prolonged survival), and nine were associated with an increased HR for death (reduced survival). Drugs associated with reduced HR included HMG-CoA reductase inhibitors (simvastatin, pravastatin, lovastatin, and atorvastatin), PDE5 inhibitors (vardenafil and sildenafil), and α-adrenergic antagonists (tamsulosin and terazosin). The medications associated with an increased HR were drugs used either in the management of clinical features of ALS associated with poor outcomes or in end-of-life care. PathFx analysis identified a complex of proteins interacting with several of the identified drugs. To our knowledge, this analysis is the largest EHR-based study for identifying drug repurposing candidates for ALS. We identified several drugs that warrant further assessment as therapeutic options in ALS, as well as a protein network complex that might serve as a therapeutic target for ALS.

Reimer, Richard J. [Stanford Univ., CA (United Sta↗
DOE OSTI2020

Bayesian inference of Stochastic reaction networks using Multifidelity Sequential Tempered Markov Chain Monte Carlo

Stochastic reaction network models are often used to explain and predict the dynamics of gene regulation in single cells.These models usually involve several parameters, such as the kinetic rates of chemical reactions, that are not directly measurable and must be inferred from experimental data. Bayesian inference provides a rigorous probabilistic frame-work for identifying these parameters by finding a posterior parameter distribution that captures their uncertainty.Traditional computational methods for solving inference problems such as Markov Chain Monte Carlo methods based on classical Metropolis-Hastings algorithm involve numerous serial evaluations of the likelihood function, which in turn requires expensive forward solutions of the chemical master equation (CME). We propose an alternate approach based on a multifidelity extension of the Sequential Tempered Markov Chain Monte Carlo (ST-MCMC) sampler. This algorithm is built upon Sequential Monte Carlo and solves the Bayesian inference problem by decomposing it into a sequence of efficiently solved subproblems that gradually increase both model fidelity and the influence of the observed data. We reformulate the finite state projection (FSP) algorithm, a well-known method for solving the CME, to produce a hierarchy of surrogate master equations to be used in this multifidelity scheme. To determine the appropriate fidelity, we introduce a novel information-theoretic criteria that seeks to extract the most information about the ultimate Bayesian posterior from each model in the hierarchy without inducing significant bias. This novel sampling scheme is tested with high performance computing resources using biologically relevant problems.

97 MATHEMATICS AND COMPUTING↗