Efficient Framework Research Articles

Efficient and Secure Aggregation Framework for Federated-Learning-Based Spectrum Sharing

Efficient and Secure Aggregation Framework for Federated-Learning-Based Spectrum Sharing

M3DIS - A grid of 3D radiation-hydrodynamics stellar atmosphere models for stellar surveys. I. Procedure, validation, and the Sun

Large-scale stellar surveys, such as SDSS-V, 4MOST, WEAVE, and PLATO, require accurate atmospheric models and synthetic spectra of stars for accurate analyses of fundamental stellar parameters and chemical abundances. The primary goal of our work is to develop a new approach to solve radiation-hydrodynamics (RHD) and generate model stellar spectra in a self-consistent and highly efficient framework. We build upon the Copenhagen legacy RHD code, the MULTI3D non-local thermodynamic equilibrium (NLTE) code, and the DISPATCH high-performance framework. The new approach allows us to calculate 3D RHD models of stellar atmospheres on timescales of a few thousand CPU hours and to perform subsequent spectrum synthesis in local thermodynamic equilibrium (LTE) or NLTE for the desired physical conditions within the parameter space of FGK-type stars. We compare the 3D RHD solar model with other available models and validate its performance against solar observations, including the centre-to-limb variation of intensities and key solar diagnostic lines of H and Fe. We show that the performance of the new code allows to overcome the main bottleneck in 3D NLTE spectroscopy and enables calculations of multi-dimensional grids of synthetic stellar observables for comparison with modern astronomical observations.

Knowledge-Enriched Prompt for Low-Resource Named Entity Recognition

Named Entity Recognition (NER) in low-resource settings aims to identify and categorize entities in a sentence with limited labeled data. Although prompt-based methods have succeeded in low-resource perspectives, challenges persist in effectively harnessing information and optimizing computational efficiency. In this work, we present a novel prompt-based method to enhance low-resource NER without exhaustive template tuning. First, we construct knowledge-enriched prompts by integrating representative entities and background information to provide informative supervision tailored to each entity type. Then, we introduce an efficient reverse generative framework inspired by question answering (QA), which avoids redundant computations. Finally, we reduce costs by generating entities from their types while retaining model reasoning capacity. Experiment results demonstrate that our method outperforms other baselines on three datasets under few-shot settings.

Low-rank reconstruction for simultaneous double half-echo 23Na and undersampled 23Na multi-quantum coherences MRI.

To develop a new sequence to simultaneously acquire Cartesian sodium (23Na) MRI and accelerated Cartesian single (SQ) and triple quantum (TQ) sodium MRI of in vivo human brain at 7 T by leveraging two dedicated low-rank reconstruction frameworks. The Double Half-Echo technique enables short echo time Cartesian 23Na MRI and acquires two k-space halves, reconstructed by a low-rank coupling constraint. Additionally, three-dimensional (3D) 23Na Multi-Quantum Coherences (MQC) MRI requires multi-echo sampling paired with phase-cycling, exhibiting a redundant multidimensional space. Simultaneous Autocalibrating and k-Space Estimation (SAKE) were used to reconstruct highly undersampled 23Na MQC MRI. Reconstruction performance was assessed against five-dimensional (5D) CS, evaluating structural similarity index (SSIM), root mean squared error (RMSE), signal-to-noise ratio (SNR), and quantification of tissue sodium concentration and TQ/SQ ratio in silico, in vitro, and in vivo. The proposed sequence enabled the simultaneous acquisition of fully sampled 23Na MRI while leveraging prospective undersampling for 23Na MQC MRI. SAKE improved TQ image reconstruction regarding SSIM by 6% and reduced RMSE by 35% compared to 5D CS in vivo. Thanks to prospective undersampling, the spatial resolution of 23Na MQC MRI was enhanced from mm3 to mm3 while reducing acquisition time from min to min. The proposed sequence, coupled with low-rank reconstructions, provides an efficient framework for comprehensive whole-brain sodium MRI, combining TSC, T2*, and TQ/SQ ratio estimations. Additionally, low-rank matrix completion enables the reconstruction of highly undersampled 23Na MQC MRI, allowing for accelerated acquisition or enhanced spatial resolution.

Enhancing Autonomous Underwater Vehicle Decision Making through Intelligent Task Planning and Behavior Tree Optimization

The expansion of underwater scenarios and missions highlights the crucial need for autonomous underwater vehicles (AUVs) to make informed decisions. Therefore, developing an efficient decision-making framework is vital to enhance productivity in executing complex tasks within tight time constraints. This paper delves into task planning and reconstruction within the AUV control decision system to enable intelligent completion of intricate underwater tasks. Behavior trees (BTs) offer a structured approach to organizing the switching structure of a hybrid dynamical system (HDS), originally introduced in the computer game programming community. In this research, an intelligent search algorithm, MCTS-QPSO (Monte Carlo tree search and quantum particle swarm optimization), is proposed to bolster the AUV’s capacity in planning complex task decision control systems. This algorithm tackles the issue of the time-consuming manual design of control systems by effectively integrating BTs. By assessing a predefined set of subtasks and actions in tandem with the complex task scenario, a reward function is formulated for MCTS to pinpoint the optimal subtree set. The QPSO algorithm is then leveraged for subtree integration, treating it as an optimal path search problem from the root node to the leaf node. This process optimizes the search subtree, thereby enhancing the robustness and security of the control architecture. To expedite search speed and algorithm convergence, this paper recommends reducing the search space by pre-grouping conditions and states within the behavior tree. The efficacy and superiority of the proposed algorithm are validated through security and timeliness evaluations of the BT, along with comparisons with other algorithms for automatic AUV decision control behavior tree design. Ultimately, the effectiveness and superiority of the proposed algorithm are corroborated through simulations on a multi-AUV complex task platform, showcasing its practical applicability and efficiency in real-world underwater scenarios.

Sparsify dynamically expandable network via variational dropout

This paper develops a new method for lifelong learning referred to as Sparsify Dynamically Expandable Network (SDEN) via Variational Dropout, which explores a sparse model while preserving the performance. Dynamically Expandable Network (DEN) can learn a sequence of tasks via performing network retraining, network expansion by adding only the necessary neurons, and network split to effectively prevent semantic drift in an online manner. To overcome point estimation of parameters in DEN, Bayesian Compression for DEN is developed under the Bayesian framework. However, this method demands more time for model training and testing. To improve the model efficiency, we propose SDEN under the efficient sparse learning framework. We validate our SDEN in the lifelong learning scenarios with multiple frequently used benchmarks, on which it can obtain comparable classification accuracy, and less training and testing time compared with the comparison methods. Furthermore, our method can also learn a more sparse network structure which means fewer network parameters.

Comparative Analysis of Organic and Chemical Pesticides: Impacts on Crop Health and Environmental Sustainability

Pesticides are indispensable tools in modern agriculture, aiding in the protection of crops against pests and diseases. Nevertheless, the adverse effects of chemical pesticides on human health and the environment have catalyzed a shift towards organic alternatives. This review critically assesses the use of both organic and chemical pesticides, evaluating their respective impacts on crop health and environmental sustainability. We delve into aspects such as efficiency, safety, environmental footprint, and regulatory frameworks surrounding these two types of pesticides, drawing insights from recent research. By synthesizing current knowledge, our aim is to provide a comprehensive understanding of the advantages and limitations of organic and chemical pesticides, empowering stakeholders to make informed decisions towards sustainable agricultural practices.

Automated marine litter investigation for underwater images using a zero-shot pipeline

Accurate and automated identification of marine litter on the seafloor is crucial due to its detrimental effects on marine ecosystems. While advancements in underwater imaging have facilitated this task, the significant human involvement required in traditional approaches necessitates the development of more efficient and cost-effective solutions. This study presents an efficient zero-shot segmentation framework based on Segment-Anything (SAM) guided by Interpretable Contrastive Language–Image Pre-training (iCLIP) for identifying and segmenting eight common seafloor litter categories in realistic underwater environments without model training. The framework supports prompt input by design, which allows it to transfer its zero-shot capabilities to new types of marine litter. To further improve the framework’s performance, two additional components were incorporated: an underwater image enhancement model that addresses the degraded image quality common in underwater environments, and a mask post-processing algorithm that reduces noise masks generated by the framework. The recorded mean intersection over union (mIOU) of 69.9% on the testing dataset suggested that zero-shot approaches have the potential to become a valuable technique for automatically detecting marine litter during surveys and enabling continuous and accurate litter monitoring.

Drone-Delivery Network for Opioid Overdose: Nonlinear Integer Queueing-Optimization Models and Methods

This study proposes an emergency network design model that uses a fleet of drones to deliver naloxone in response to opioid overdoses. The network is represented as a collection of M/G/K queueing systems with variable capacity and service time modelled as a decision-dependent random variable. The model is a complex queuing-based optimization problem which locates drone bases and dispatches drones to opioid incidents. The authors devise an efficient solution framework in which they linearize the multiple nonlinearities (fractional, polynomial, exponential, factorial terms), derive an equivalent mixed-integer linear reformulation, and design an outer approximation branch-and-cut algorithm. The authors demonstrate the generalizablity of the approach to any problem minimizing the response time of M/G/K queueing systems with unknown capacity. Tests on Virginia Beach data reveal that drones can decrease response time by 82%, increase survival chance by more than 273%, save up to 33 additional lives annually, and provide up to 279 additional quality-adjusted life years.

Fast Fractional Fourier Transform-Aided Novel Graphical Approach for EEG Alcoholism Detection

Given its detrimental effect on the brain, alcoholism is a severe disorder that can produce a variety of cognitive, emotional, and behavioral issues. Alcoholism is typically diagnosed using the CAGE assessment approach, which has drawbacks such as being lengthy, prone to mistakes, and biased. To overcome these issues, this paper introduces a novel paradigm for identifying alcoholism by employing electroencephalogram (EEG) signals. The proposed framework is divided into various steps. To begin, interference and artifacts in the EEG data are removed using a multiscale principal component analysis procedure. This cleaning procedure contributes to information quality improvement. Second, an innovative graphical technique based on fast fractional Fourier transform coefficients is devised to visualize the chaotic character and complexities of the EEG signals. This elucidates the properties of regular and alcoholic EEG signals. Third, thirty-four graphical features are extracted to interpret the EEG signals’ haphazard behavior and differentiate between regular and alcoholic trends. Fourth, we propose an ensembled feature selection method for obtaining an effective and reliable feature group. Following that, we study many neural network classifiers to choose the optimal classifier for building an efficient framework. The experimental findings show that the suggested method obtains the best classification performance by employing a recurrent neural network (RNN), with 97.5% accuracy, 96.7% sensitivity, and 98.3% specificity for the sixteen selected features. The proposed framework can aid physicians, businesses, and product designers to develop a real-time system.

An Efficient Framework for Localization Based Optimized Opportunistic Routing Protocol in Underwater Acoustic Sensor Networks

An Efficient Framework for Localization Based Optimized Opportunistic Routing Protocol in Underwater Acoustic Sensor Networks

Securing File Storage by Hybrid Cryptography

The proliferation of digital data in modern times necessitates robust security measures to safeguard confidentiality and integrity. This paper introduces a "File Storage System using Hybrid Cryptography," a pivotal solution addressing the pressing need for secure file storage and retrieval. Hybrid cryptography merges symmetric and asymmetric encryption techniques, ensuring a resilient and reliable data protection mechanism. This system leverages the strengths of symmetric encryption's speed and efficiency while integrating the key management advantages of asymmetric encryption. Through an exploration of historical cryptographic foundations, the emergence of hybrid cryptography, and contemporary file storage challenges, this paper highlights the importance of such an approach. The proposed system's phases, from encryption to decryption, elucidate its implementation and efficacy. Future research directions are outlined, focusing on enhanced key management, performance optimization, multi-platform integration, and user-centric solutions. Challenges such as quantum computing threats, scalability, and integration with emerging technologies are discussed. The conclusion emphasizes the significance of hybrid cryptography in addressing data security concerns amidst escalating digital data volumes, offering an efficient and secure framework for data storage and sharing. The research gaps identified prompt further exploration into quantum-resistant systems, scalability optimization, user-friendly interfaces, cross-cultural legal considerations, ethical implications, sustainability, and integration with emerging technologies.

Efficient Crowd Anomaly Detection Using Sparse Feature Tracking and Neural Network

Efficient Crowd Anomaly Detection Using Sparse Feature Tracking and Neural Network

An efficient GS-RBFN framework for early prediction and classification of ad

An efficient GS-RBFN framework for early prediction and classification of ad

Design of an Iterative Method for Optimizing Solar Power Systems using Quad LSTM with IoT Integration Operations

The growing necessity for sustainable energy solutions underscores the critical need for optimizing solar power systems. Existing methodologies predominantly focus on static strategies with limited adaptability to dynamic environmental and operational conditions, often resulting in suboptimal performance and inefficiency. This research introduces a comprehensive, integrated framework employing Internet of Things (IoT) technologies alongside advanced machine learning and deep learning methodologies to enhance solar power system efficiency and reliability levels. The proposed model integrates four key components: predictive maintenance using Support Vector Machines (SVM) for enhanced anomaly detection, solar power forecasting via Quad Long Short-Term Memory (QLSTM) neural networks, dynamic load balancing through Reinforcement Learning with Deep Q-learning, and the integration with smart grids employing Decentralized Multi-Agent Systems (MAS) with Auction-Based Mechanisms. Each method is selected based on its suitability to address specific challenges within the solar power domain: SVMs for their effectiveness in high-dimensional anomaly detection, QLSTMs for their superior temporal pattern recognition in forecasting, Deep Q-learning for its adaptability in dynamic load management, and MAS for efficient decentralized energy resource coordination operations. The implementation of these methodologies demonstrates significant advancements over traditional approaches. Predictive maintenance facilitated by SVMs leads to a 20% reduction in maintenance costs, while QLSTM-based forecasting achieves a 95% accuracy rate, thereby enhancing grid management and reducing revenue losses. Moreover, reinforcement learning optimizes energy utilization, decreasing wastage and system downtime by 10% and 5% respectively. Lastly, the MAS framework promotes a 20% increase in energy trading efficiency, yielding a 10% reduction in transaction costs and bolstering grid resilience levels. This work represents a significant leap forward in solar power optimization, offering a scalable, efficient, and intelligent framework that paves the way for more sustainable and reliable energy systems. The integration of IoT with machine learning and deep learning presents a paradigm shift in renewable energy management, marking a critical step toward achieving global sustainability objectives..

Enhancements in Immediate Speech Emotion Detection: Harnessing Prosodic and Spectral Characteristics

Speech is essential to human communication for expressing and understanding feelings. Emotional speech processing has challenges with expert data sampling, dataset organization, and computational complexity in large-scale analysis. This study aims to reduce data redundancy and high dimensionality by introducing a new speech emotion recognition system. The system employs Diffusion Map to reduce dimensionality and includes Decision Trees and K-Nearest Neighbors(KNN)ensemble classifiers. These strategies are suggested to increase voice emotion recognition accuracy. Speech emotion recognition is gaining popularity in affective computing for usage in medical, industry, and academics. This project aims to provide an efficient and robust real-time emotion identification framework. In order to identify emotions using supervised machine learning models, this work makes use of paralinguistic factors such as intensity, pitch, and MFCC. In order to classify data, experimental analysis integrates prosodic and spectral information utilizing methods like Random Forest, Multilayer Perceptron, SVM, KNN, and Gaussian Naïve Bayes. Fast training times make these machine learning models excellent for real-time applications. SVM and MLP have the highest accuracy at 70.86% and 79.52%, respectively. Comparisons to benchmarks show significant improvements over earlier models.

Integrated Optimization of Hybrid Steam-Solvent Injection in Post-CHOPS Reservoirs with Consideration of Wormhole Networks and Foamy Oil Behavior

Summary For this paper, integrated techniques have been developed to optimize the performance of the hybrid steam-solvent injection processes in a depleted post-cold heavy oil production with sand (CHOPS) reservoir with consideration of wormhole networks and foamy oil behavior. After a reservoir geological model has been built and calibrated with the measured production profiles, its wormhole network is inversely determined using the newly developed pressure-gradient-based (PGB) sand failure criterion. Such a calibrated reservoir geological model is then used to maximize the net present value (NPV) of a hybrid steam-solvent injection process by selecting injection time, soaking time, production time, injection rate, steam temperature, and steam quality as the controlling variables. The genetic algorithm (GA) has been integrated with orthogonal array (OA) and Tabu search to maximize the NPV by delaying the displacement front as well as extending the reservoir life under various strategies. Considering the wormhole network and foamy oil behavior and using the NPV as the objective function, such a modified algorithm can be used to allocate and optimize the production-injection strategies of each huff ‘n’ puff (HnP) cycle in a post-CHOPS reservoir with altered porosity and increased permeability within a unified, consistent, and efficient framework.

Automated Chemical Reaction Network Generation and Its Application to Exoplanet Atmospheres

With the advent of JWST and the spectroscopic characterization of exoplanet atmospheres in unprecedented detail, there is a demand for more complete pictures of chemical and photochemical reactions and their impacts on atmospheric composition. Traditionally, building reaction networks for (exo)planetary atmospheres involves manually tracking relevant species and reactions, a time-consuming and error-prone process. This approach’s applicability is also often limited to specific conditions, making it less versatile for different planetary types (i.e., photochemical networks for Jupiters may not be directly applicable to water-rich exoplanets). We introduce an automated approach using a computer-aided chemical reaction network generator, combined with a 1D photochemical kinetic-transport model, offering significant advantages. This approach automatically selects reaction rates through a rate-based iterative algorithm and multiple refinement steps, enhancing model reliability. Also, this approach allows for the efficient simulation of diverse chemical environments, from hydrogen to water, carbon dioxide, and nitrogen-dominated atmospheres. Using WASP-39b and WASP-80b as examples, we demonstrate our approach’s effectiveness, showing good agreement with recent JWST data. Our WASP-39b model aligns with prior studies and JWST observations, capturing photochemically produced sulfur dioxide. The WASP-80b model reveals an atmosphere influenced by deep-interior thermochemistry and vertical mixing, consistent with JWST NIRCam observations. Furthermore, our model identifies a novel initial step for the N2–NH3–HCN pathway that enhances the efficiency of the conversion in high-temperature/high-pressure environments. This automated chemical network generation offers a novel, efficient, and precise framework for studying exoplanetary atmospheres, marking a significant advancement over traditional modeling techniques.

A Bayesian optimization framework for the control of combustion instability of a bluff-body stabilized combustor

Control of combustion instability for a realistic gas-turbine combustor is challenging. This work aims to establish an efficient numerical framework for optimization to improve the combustion stability of a bluff-body combustor. Large eddy simulations of the spray combustion process are conducted, and the experimental measurements are used to evaluate the numerical accuracy of the baseline case. The air preheating temperature, the Sauter mean diameter of fuel droplets, and the location of liquid fuel injection are regarded as input variables. The root mean square of pressure amplitude is regarded as an optimization objective. The Bayesian optimization framework is proposed that includes the sampling process, surrogate model, acquisition function, and genetic algorithm optimizer processes. It is found that PRMS can be reduced by 64% for the optimized case compared to the baseline case using only 17 sample evaluations. This work is promising as it provides an effective optimization framework for the development of next-generation gas-turbine combustors.

Development of RAMDOE: a new method for rapidly ranking alternatives with supplementary options and considering changes in criteria values

This paper delves into the development and validation of the RAMDOE method, a pioneering approach in multi-criteria decision making (MCDM) that seamlessly integrates the root assessment method (RAM) and design of experiments (DOE) techniques, addressing the inflexibility of traditional MCDM methods in accommodating adjustments in criteria ranges and the addition of new alternatives without necessitating a complete overhaul of the decision framework. Through empirical analysis, the study demonstrates the RAMDOE method's remarkable efficacy in precisely ranking alternatives, as illustrated through a practical case study focused on the selection of a supplier from a pool of seven candidates. One of the most notable aspects of the RAMDOE method lies in its capacity to formulate a regression equation that accurately captures the intricate relationship between alternative scores and criteria values, enabling decision-makers to seamlessly integrate new alternatives into the decision-making process without the cumbersome task of recalibration, thereby distinguishing it from conventional MCDM techniques such as TOPSIS (technique for order of preference by similarity to ideal solution), COPRAS (complex proportional assessment), MOORA (multiobjective optimization on the basis of ratio analysis), EDAS (evaluation based on distance from average solution) and CODAS (combinative distance-based assessment). The practical implications of these findings are profound, offering decision-makers across various domains a more efficient and adaptable framework to navigate complex decision scenarios. Particularly in contexts like supplier selection, where criteria ranges may vary significantly, the RAMDOE method provides decision-makers with a robust toolset to make informed decisions, presenting a promising avenue for addressing the dynamic nature of decision-making environments and enhancing the overall robustness and flexibility of MCDM processes in real-world applications