Computational Cost Research Articles

Asymptotic Efficiency of Joint Estimator Relative to Two-Stage Estimator Under Misspecified Likelihoods

Abstract The two-stage estimator is often more tractable when there are nuisance parameters that can be separately estimated and plugged into an objective function. The joint estimator tends to bear the higher computational cost since it estimates all parameters in one stage by optimizing the sum of objective functions used in the two stages. It is well-known that the joint estimator is asymptotically more efficient than the two-stage estimator if the objective function is the true log-likelihood. When the objective function is not the true log-likelihood, I show that the relative asymptotic efficiency of the joint estimator still holds under a finite number of testable moment conditions. The implications of the main result on models based on quasi-limited information likelihoods are discussed.

Hierarchical Identity-Based Authenticated Encryption with Keyword Search over encrypted cloud data

With the rapid development of cloud computing technology, cloud storage services are becoming more and more mature. However, the storage of sensitive data on remote servers poses privacy risks and is presently a source of concern. Searchable Encryption (SE) is an effective method for protecting sensitive data while preserving server-side searchability. Hierarchical Public key Encryption with Keyword Search (HPEKS), a new variant of SE, allows users with higher access permission to search over encrypted data sent to lower-level users. To the best of our knowledge, there exist only four HPEKS schemes in the literature. Two of them are in traditional public-key setting, and the remaining ones are identity-based public key cryptosystems. Unfortunately, all of the four existing HPEKS schemes are vulnerable against inside Keyword Guessing Attacks (KGAs). Moreover, all of the existing HPEKS schemes are based on the computationally expensive bilinear pairing operation which dramatically increases the computational costs. To overcome these issues, in this paper, we introduce the notion of Hierarchical Identity-Based Authenticated Encryption with Keyword Search (HIBAEKS). We formulate a security model for HIBAEKS and propose an efficient pairing-free HIBAEKS scheme. We then prove that the proposed HIBAEKS scheme is secure under the defined security model and is resistant against KGAs. Finally, we compare our proposed scheme with related constructions regarding security requirements, computational and communication costs to indicate the overall superiority of our proposed scheme.

In Silico Chemical Experiments in the Age of AI: From Quantum Chemistry to Machine Learning and Back.

Computational chemistry is an indispensable tool for understanding molecules and predicting chemical properties. However, traditional computational methods face significant challenges due to the difficulty of solving the Schrödinger equationsand the increasing computational cost with the size of the molecular system. In response, there has been a surge of interest in leveraging artificial intelligence (AI) and machine learning (ML) techniques to in silico experiments. Integrating AI and ML into computational chemistry increases the scalability and speed of the exploration of chemical space. However, challenges remain, particularly regarding the reproducibility and transferability of ML models. This review highlights the evolution of ML in learning from, complementing, or replacing traditional computational chemistry for energy and property predictions. Starting from models trained entirely on numerical data, a journey set forth toward the ideal model incorporating or learning the physical laws of quantum mechanics. This paper also reviews existing computational methods and ML models and their intertwining, outlines a roadmap for future research, and identifies areas for improvement and innovation. Ultimately, the goal is to develop AI architectures capable of predicting accurate and transferable solutions to the Schrödinger equation, thereby revolutionizing in silico experiments within chemistry and materialsscience.

Mixture-of-experts and semantic-guided network for brain tumor segmentation with missing MRI modalities.

Accurate brain tumor segmentation with multi-modal MRI images is crucial, but missing modalities in clinical practice often reduce accuracy. The aim of this study is to propose a mixture-of-experts and semantic-guided network to tackle the issue of missing modalities in brain tumor segmentation. We introduce a transformer-based encoder with novel mixture-of-experts blocks. In each block, four modality experts aim for modality-specific feature learning. Learnable modality embeddings are employed to alleviate the negative effect of missing modalities. We also introduce a decoder guided by semantic information, designed to pay higher attention to various tumor regions. Finally, we conduct extensive comparison experiments with other models as well as ablation experiments to validate the performance of the proposed model on the BraTS2018 dataset. The proposed model can accurately segment brain tumor sub-regions even with missing modalities. It achieves an average Dice score of 0.81 for the whole tumor, 0.66 for the tumor core, and 0.52 for the enhanced tumor across the 15 modality combinations, achieving top or near-top results in most cases, while also exhibiting a lower computational cost. Our mixture-of-experts and sematic-guided network achieves accurate and reliable brain tumor segmentation results with missing modalities, indicating its significant potential for clinical applications. Our source code is already available at https://github.com/MaggieLSY/MESG-Net .

Elastic reverse time migration based on nested triangular mesh in 2D isotropic media

Abstract Reverse time migration (RTM) based on two-way wave equation is an important imaging tool, which has been widely used in exploration geophysics. It offers distinct advantages in handling complex geological structures. Moreover, elastic reverse time migration (ERTM) can produce more physically meaningful images for subsurface and has received increasing attention in recent years. However, rugged topography presents challenges in the reconstruction of subsurface structures for ERTM. To overcome this problem, we introduce the grid method for topography ERTM, which is an unstructured mesh based algorithm. However, the conventional gird method for RTM is implementing by discretizing model into small meshes, leading to high computational cost. To address this problem, we introduce the grid method based on nested unstructured meshes to implement ERTM with complex surface topography. Low-order small meshes, due to their fine scale, can accurately depict complex surface topography. However, their generation process is relatively time-consuming. In contrast, high-order large meshes contain many similar small meshes internally, which can be processed in parallel, resulting in higher computational efficiency. Therefore, we propose the following strategy: for the receiver wavefield, several layers on surface discreted by the small meshes are used to place receivers and remaining layers discreted by the large meshes to improve computational efficiency. For the source wavefield, only the large meshes are used because the spacing of shot is relatively large. In this way, we ensure the accurate description of complex surface topography for the implementation of ERTM with high computational efficiency. Finally, the effectiveness of our method is validated through experiments on three models with complex topography.

Internal temperature estimation for lithium-ion batteries through distributed equivalent circuit network model

Lithium-ion cells experience significant internal thermal gradients during operation, with a direct impact on their safety, performance, cost and lifetime. The estimation of the internal temperature of cells is therefore particularly important. In this work, a 3D distributed electro-thermal model for internal temperature estimation is developed for a cylindrical cell (LG M50T, NMC811). The model is parameterized and comprehensively validated against experimental data for 21700 cells, including direct core temperature measurements. Multiple types of electrical load are considered, including constant current discharge, pulse discharge, drive cycle and instant discharge/charge switching. The developed model is used to estimate core temperature based on surface temperature measurement. The predictions are shown to have good accuracy at relatively low computational cost. We show that the widely adopted two-node lumped thermal estimation model is increasingly inaccurate for more aggressive discharges, when thermal gradients become higher. Compared to the standard two-node model, the distributed equivalent circuit network model predicts the effects of detailed internal cell structure (electrode, current collector, metal can and tab) and distributed internal heat generation. The results are of immediate interest to cell manufacturers and battery pack designers, while the modelling and parameterization framework is a useful tool for energy storage systems design.

Reducing the Runtime of Fault-Tolerant Quantum Simulations in Chemistry through Symmetry-Compressed Double Factorization.

Quantum phase estimation based on qubitization is the state-of-the-art fault-tolerant quantum algorithm for computing ground-state energies in chemical applications. In this context, the 1-norm of the Hamiltonian plays a fundamental role in determining the total number of required iterations and also the overall computational cost. In this work, we introduce the symmetry-compressed double factorization (SCDF) approach, which combines a CDF of the Hamiltonian with the symmetry shift technique, significantly reducing the 1-norm value. The effectiveness of this approach is demonstrated numerically by considering various benchmark systems, including the FeMoco molecule, cytochrome P450, and hydrogen chains of different sizes. To compare the efficiency of SCDF to other methods in absolute terms, we estimate Toffoli gate requirements, which dominate the execution time on fault-tolerant quantum computers. For the systems considered here, SCDF leads to a sizable reduction of the Toffoli gate count in comparison to other variants of DF or even tensor hypercontraction, which is usually regarded as the most efficient approach for qubitization.

Hyperspectral Image Classification Based on Two-Branch Multiscale Spatial Spectral Feature Fusion with Self-Attention Mechanisms

In recent years, the use of deep neural network in effective network feature extraction and the design of efficient and high-precision hyperspectral image classification algorithms has gradually become a research hotspot for scholars. However, due to the difficulty of obtaining hyperspectral images and the high cost of annotation, the training samples are very limited. In order to cope with the small sample problem, researchers often deepen the network model and use the attention mechanism to extract features; however, as the network model continues to deepen, the gradient disappears, the feature extraction ability is insufficient, and the computational cost is high. Therefore, how to make full use of the spectral and spatial information in limited samples has gradually become a difficult problem. In order to cope with such problems, this paper proposes two-branch multiscale spatial–spectral feature aggregation with a self-attention mechanism for a hyperspectral image classification model (FHDANet); the model constructs a dense two-branch pyramid structure, which can achieve the high efficiency extraction of joint spatial–spectral feature information and spectral feature information, reduce feature loss to a large extent, and strengthen the model’s ability to extract contextual information. A channel–space attention module, ECBAM, is proposed, which greatly improves the extraction ability of the model for salient features, and a spatial information extraction module based on the deep feature fusion strategy HLDFF is proposed, which fully strengthens feature reusability and mitigates the feature loss problem brought about by the deepening of the model. Compared with five hyperspectral image classification algorithms, SVM, SSRN, A2S2K-ResNet, HyBridSN, SSDGL, RSSGL and LANet, this method significantly improves the classification performance on four representative datasets. Experiments have demonstrated that FHDANet can better extract and utilise the spatial and spectral information in hyperspectral images with excellent classification performance under small sample conditions.

A Blockchain-Based Electronic Health Record (EHR) System for Edge Computing Enhancing Security and Cost Efficiency

Blockchain technology offers unique features, such as transparency, the immutability of data, and the capacity to establish trust without a central authority. Such characteristics can be leveraged to support the collaboration among several different software systems operating within the healthcare ecosystem, while ensuring data integrity and make electronic health records (EHRs) more easily accessible. To provide a solution based on blockchain technology, this paper has evaluated the main issues that arise when large amounts of data are expected, i.e., mainly cost and performance. A balanced approach that maximizes the benefits and mitigates the constraints of the blockchain has been designed. The proposed decentralized application (dApp) architecture employs a hybrid storage strategy that involves storing medical records locally, on users’ devices, while utilizing blockchain to manage an index of these data. The dApp clients facilitate interactions among participants, leveraging a smart contract to enable patients to set authorization policies, thereby ensuring that only designated healthcare providers and authorized entities have access to specific medical records. The blockchain data-immutability property is used to validate data stored externally. This solution significantly reduces the costs related to the utilization of the blockchain, while retaining its advantages, and improves performance, since the majority of data are available off-chain.

Quantum Quality with Classical Cost: Ab Initio Nonadiabatic Dynamics Simulations Using the Mapping Approach to Surface Hopping.

Nonadiabatic dynamics methods are an essential tool for investigating photochemical processes. In the context of employing first-principles electronic structure techniques, such simulations can be carried out in a practical manner using semiclassical trajectory-based methods or wave packet approaches. While all approaches applicable to first-principles simulations are necessarily approximate, it is commonly thought that wave packet approaches offer inherent advantages over their semiclassical counterparts in terms of accuracy and that this trait simply comes at a higher computational cost. Here we demonstrate that the mapping approach to surface hopping (MASH), a recently introduced trajectory-based nonadiabatic dynamics method, can be efficiently applied in tandem with ab initio electronic structure. Our results even suggest that MASH may provide more accurate results than on-the-fly wave packet techniques, all at a much lower computational cost.

DTP: learning to estimate full-body pose in real-time from sparse VR sensor measurements

For virtual reality (VR) applications, estimating full-body pose in real-time is becoming increasingly popular. Previous works have reconstructed full-body motion in real time from an HTC VIVE headset and five VIVE Tracker measurements by solving the inverse kinematics (IK) problem. However, an IK solver may yield unnatural poses and shaky motion. This paper introduces Deep Tracker poser (DTP): a method for real-time full-body pose estimation in VR. This task is difficult due to the ambiguous mapping from the sparse measurements to full-body pose. The data obtained from VR sensors is calibrated, normalized and fed into the deep neural networks (DNN). To learn from sufficient data, we propose synthesizing a VR sensor dataset called AMASS-VR from the AMASS, a collection of various motion capture datasets. Furthermore, feet tracking loss is a common problem of VIVE Tracker. To improve the accuracy and robustness of DTP to the occlusion noise, we simulate the occlusion noise by Gaussian random noise. Then we synthesize an occlusion dataset AMASS-OCC and fine-tune DTP on that. We evaluate DTP by comparing with other popular methods in terms of the accuracy and computational cost. The results indicate that DTP outperforms others in terms of the positional error (1.04 cm) and rotational error (4.22 °). The quantitative and qualitative results show that DTP reconstructs accurate and natural full-body pose even under serious feet occlusion, which indicates the superiority of the DTP in modelling the mapping from sparse joint data to the full-body pose.

ER‐Mapping: An extrinsic robust coloured mapping system using residual evaluation and selection

AbstractThe colour‐enhanced point cloud map is increasingly being employed in fields such as robotics, 3D reconstruction and virtual reality. The authors propose ER‐Mapping (Extrinsic Robust coloured Mapping system using residual evaluation and selection). ER‐Mapping consists of two components: the simultaneous localisation and mapping (SLAM) subsystem and the colouring subsystem. The SLAM subsystem reconstructs the geometric structure, where it employs a dynamic threshold‐based residual selection in LiDAR‐inertial odometry to improve mapping accuracy. On the other hand, the colouring subsystem focuses on recovering texture information from input images and innovatively utilises 3D–2D feature selection and optimisation methods, eliminating the need for strict hardware time synchronisation and highly accurate extrinsic parameters. Experiments were conducted in both indoor and outdoor environments. The results demonstrate that our system can enhance accuracy, reduce computational costs and achieve extrinsic robustness.

Block Spectral Method Analysis for Turbomachinery Applications

Abstract The block-wise spectral reconstruction of turbomachinery flows has been successfully used in the past to reduce the computational cost of full-annulus turbomachinery simulations. The rationale of the method is based on an explicit representation of the long-wavelengths of the problem using a Fourier expansion of the blocks, which in turn contain the short-wavelength content of the spectrum. This work analyzes the accuracy of this approach and shows that this approximation retains not only the explicit Fourier representation of the long wavelengths and finite volume approximation of the short wavelengths but their interaction as well. This interaction due to the scattering associated with the nonuniformity of the flow is referred to as Tyler–Sofrin modes by turbomachinery practitioners and is highly relevant for many practical applications. The solution’s harmonic content is illustrated by resorting to the linear wave equation with varying propagation speed and the NASA rotor 67 fan.

A shortest‐path‐aided fast‐sweeping method to improve the accuracy of traveltime calculation in vertically transverse isotropic media

AbstractThe high accuracy and efficiency of traveltime calculation are critical in seismic tomography, migration, static corrections, source locations and anisotropic parameter estimation. The fast‐sweeping method is an efficient upwind finite‐difference approach for solving the eikonal equation. However, the fast‐sweeping method is accurate only along the axis directions. In two‐dimensional or higher dimensional cases, the accuracy is severely decreased in the diagonal directions due to the numerical errors in these directions. These similar numerical errors also arose in higher order fast‐sweeping method and anisotropic fast‐sweeping method. To improve the accuracy of traveltime calculation in two‐dimensional or higher dimensional space, a shortest‐path‐aided fast‐sweeping method is proposed. The shortest‐path‐aided solution is embedded into the sweeping process of the standard fast‐sweeping method to improve the traveltime accuracy in the diagonal directions. Shortest‐path‐aided fast‐sweeping method is very easy to implement nearly without additional computational cost and memory consumption. Furthermore, this method is easy to extend from two‐dimensional to higher dimensional, from low‐order to higher‐order and from isotropic to anisotropic cases.

Simple model to incorporate statistical noise based on a modified hodgkin-huxley approach for external electrical field driven neural responses.

Noise activity is known to affect neural networks, enhance the system response to weak external signals, and lead to stochastic resonance phenomenon that can effectively amplify signals in nonlinear systems. In most treatments, channel noise has been modeled based on multi-state Markov descriptions or the use stochastic differential equation models. Here we probe a computationally simple approach based on a minor modification of the traditional Hodgkin-Huxley approach to embed noise in neural response. Results obtained from numerous simulations with different excitation frequencies and noise amplitudes for the action potential firing show very good agreement with output obtained from well-established models. Furthermore, results from the Mann-Whitney U test reveal a statistically insignificant difference. The distribution of the time interval between successive potential spikes obtained from this simple approach compared very well with the results of complicated Pu-Thomas type methods at much reduced computational cost. This present method could also possibly be applied to the analysis of spatial variations and/or differences in characteristics of random incident electromagnetic signals.

Can physics-informed neural networks beat the finite element method?

Abstract Partial differential equations play a fundamental role in the mathematical modelling of many processes and systems in physical, biological and other sciences. To simulate such processes and systems, the solutions of PDEs often need to be approximated numerically. The finite element method, for instance, is a usual standard methodology to do so. The recent success of deep neural networks at various approximation tasks has motivated their use in the numerical solution of PDEs. These so-called physics-informed neural networks and their variants have shown to be able to successfully approximate a large range of partial differential equations. So far, physics-informed neural networks and the finite element method have mainly been studied in isolation of each other. In this work, we compare the methodologies in a systematic computational study. Indeed, we employ both methods to numerically solve various linear and nonlinear partial differential equations: Poisson in 1D, 2D, and 3D, Allen–Cahn in 1D, semilinear Schrödinger in 1D and 2D. We then compare computational costs and approximation accuracies. In terms of solution time and accuracy, physics-informed neural networks have not been able to outperform the finite element method in our study. In some experiments, they were faster at evaluating the solved PDE.

Coordinated Control of the Onboard and Wayside Energy Storage System of an Urban Rail Train Based on Rule Mining

AbstractThere are three major challenges to the broad implementation of energy storage systems (ESSs) in urban rail transit: maximizing the absorption of regenerative braking power, enabling online global optimal control, and ensuring algorithm portability. To address these problems, a coordinated control framework between onboard and wayside ESSs is proposed in this study, and the related control strategy is obtained by transforming the global optimization problem into a combination of rule-based and local optimization problems. The design process of the strategy is divided into three parts: offline optimization, rule extraction, and local optimization. First, a genetic algorithm is used to optimize the charge/discharge threshold curves under typical operational conditions. Then, the rules of offline optimization results are obtained from an expert system, and the nonlinear rules are mined with local optimization methods. Finally, a coordinated strategy is presented based on the outcomes of the three parts. A power hardware-in-the-loop (PHIL) experimental platform based on RTLAB is built, and the above coordination strategy is verified based on data for the Beijing Metro Batong Line. The experimental results show that the strategy can effectively improve the energy-saving rate and reduce the regeneration failure rate substantially, with the effect being close to the offline global optimal solution. The algorithm proposed in this paper achieves near global optimal energy-saving optimization results with lower computational costs, and has strong portability, providing a good solution for the large-scale application of onboard and wayside energy storage coordination control.

Estimating Forest Aboveground Biomass Using a Combination of Geographical Random Forest and Empirical Bayesian Kriging Models

Accurately estimating forest aboveground biomass (AGB) is imperative for comprehending carbon cycling, calculating carbon budgets, and formulating sustainable forest management plans. Currently, random forest (RF) and other machine learning models are widely used to estimate forest AGB, as they can effectively handle nonlinear relationships. However, by constructing a global model using all the samples collected from a study area, these models fail to account for the spatial heterogeneity in the AGB and cannot correct the prediction biases, thereby constraining the estimation accuracy. To overcome these limitations, we proposed a novel approach termed geographical random forest and empirical Bayesian kriging (GRFEBK). This hybrid model combines the localized modeling capability of geographical random forest (GRF) with the bias correction strength of empirical Bayesian kriging (EBK). GRF adapts RF to account for the spatial heterogeneity of the AGB, while EBK utilizes the spatial autocorrelation of residuals to correct the prediction deviations. This study was conducted in Hainan Island, utilizing spectral bands, vegetation indices, tasseled cap components derived from Landsat-8 imagery, backscattering coefficients from ALOS-2 synthetic aperture radar, topographic features, and the forest canopy height as the explanatory variables. A total of 195 forest aboveground biomass (AGB) samples were collected for modeling and assessing the predictive accuracy. The results demonstrate that, among the tested models, including GRFEBK, RF, support vector machine (SVM), k-nearest neighbor (KNN), geographically weighted regression (GWR), GRF, and EBK, GRFEBK attains the highest R2 (0.78) and the lowest RMSE (36.04 Mg/ha) and RRMSE (22.87%), significantly outperforming the conventional models and using GRF or EBK alone. These results demonstrate that by accounting for local non-stationarity in AGB and correcting prediction biases, GRFEBK achieves significantly higher accuracy than conventional RF and other models. While the results are promising, the computational cost of GRFEBK and its performance under varying geographical conditions warrant further investigation at larger scales to assess its broader applicability. Nevertheless, GRFEBK provides an innovative and more reliable approach for accurate forest AGB estimation with great potential to support global forest resource monitoring.

Calibrated confidence learning for large-scale real-time crash and severity prediction

Real-time crash and severity prediction is a complex task, and there is no existing framework to predict crash likelihood and severity together. Creating such a framework poses numerous challenges, particularly not independent and identically distributed (non-IID) data, large model sizes with high computational costs, missing data, sensitivity vs. false alarm rate (FAR) trade-offs, and real-world deployment strategies. This study introduces a novel modeling technique to address these challenges and develops a deployable real-world framework. We used extensive real-time traffic and weather data to develop a crash likelihood prediction modeling prototype, leveraging our preliminary work of spatial ensemble modeling. Next, we equipped this spatial ensemble model with local model regularization to calibrate model confidence training. The investigated regularizations include weight decay, label smoothing and knowledge distillation. Furthermore, post-calibration on model outputs was conducted to improve severity rating identification. We tested the framework to predict crashes and severity in real-time, categorizing crashes into four levels. Results were compared with benchmark models, real-world deployment mechanisms were explained, traffic safety improvement potential and sustainability aspects of the study were discussed. Modeling results demonstrated excellent performance, and fatal, severe, minor and PDO crash severities were predicted with 91.7%, 83.3%, 85.6%, and 87.7% sensitivity, respectively, and with very low FAR. Similarly, the viability of our model to predict different severity levels for specific crash types, i.e., all-crash types, rear-end crashes, and sideswipe/angle crashes, were examined, and it showed excellent performance. Our modeling technique showed great potential for reducing model size, lowering computational costs, improving sensitivity, and, most importantly, reducing FAR. Finally, the deployment strategy for the proposed crash and severity prediction technique is discussed, and its potential to predict crashes with severity levels in real-time will make a substantial contribution to tailoring specific strategies to prevent crashes.

A concept for the global assessment of tomographic resolution and uncertainty

Summary A major challenge in seismic tomography consists in quantifying and representing model resolution and uncertainty, particularly at global scales. This information is crucial for interpretations of tomographic images and their technical application in geodynamics. However, due to large computational costs, there have been only few attempts so far to coherently analyse the spatially varying resolving power for a complete set of model parameters. Here, we present a concept for an effective evaluation and global representation of the 3-D resolution information contained in a full set of averaging kernels. In our case, these kernels are constructed using the ‘Subtractive Optimally Localized Averages’ (SOLA) method, a variant of classic Backus-Gilbert inversion suitable for global tomography. Our assessment strategy incorporates the following steps: 1) a 3-D Gaussian function is fitted to each averaging kernel to measure resolution lengths in different directions; 2) we define a classification scheme for the quality of the averaging kernels based on their focus with respect to the estimated 3-D Gaussian, allowing us to reliably identify whether the inferred resolution lengths are robust. This strategy is not restricted to SOLA inversions, but can, for example, be applied in all cases where point-spread functions are computed in other tomographic frameworks. Together with model uncertainty estimates that are derived from error propagation in the SOLA method, our concept reveals at which locations, resolution lengths and interpretations of model values are actually meaningful. We finally illustrate how the complete information from our analysis can be used to calibrate the SOLA inversion parameters —locally tunable target resolution kernels and trade-off parameters— without the need for visual inspection of the individual resulting averaging kernels. Instead, our global representations provide a tool for designing tomographic models with specific resolution-uncertainty properties that are useful in geodynamic applications, especially for linking seismic inversions to models of mantle flow.