The new measurement scheme of IP performance metrics is for the mobile network in heterogeneous wireless network environment. In the proposed scheme, when Mobile Nodes （MNs） inside the mobile network needs to under...The new measurement scheme of IP performance metrics is for the mobile network in heterogeneous wireless network environment. In the proposed scheme, when Mobile Nodes （MNs） inside the mobile network needs to understand the condition of multiple comrmunicatinn paths outside the mobile netwtrk, they can get IP performance metrics, such as delay, jitter, bandwidth, packet loss, etc., irrespective of the preserre or absence of measurement functionality. At the same time, the proposed scheme dees not require the MN to he involved in measuring IP performance metrice. The Multihomed Mobile Router （MMR） with heterogeneons wireless interfaces measures IP performance metrics on behalf of the MNs inside the mobile network. Then, MNs can get measured IP perfonmnce metries from the MMR using L3 messages. The proposed scheme can reduce burden and power consumption of MNs with limited resource and batty power since MNs don＇ t measure IP performance metrics directly. In addition, it can reduce considerably traffic overhead over wireless links on multiple measurement paths since signaling messages and injeeted testing traffic are reduced.展开更多
Cross entropy is a measure in machine learning and deep learning that assesses the difference between predicted and actual probability distributions. In this study, we propose cross entropy as a performance evaluation...Cross entropy is a measure in machine learning and deep learning that assesses the difference between predicted and actual probability distributions. In this study, we propose cross entropy as a performance evaluation metric for image classifier models and apply it to the CT image classification of lung cancer. A convolutional neural network is employed as the deep neural network (DNN) image classifier, with the residual network (ResNet) 50 chosen as the DNN archi-tecture. The image data used comprise a lung CT image set. Two classification models are built from datasets with varying amounts of data, and lung cancer is categorized into four classes using 10-fold cross-validation. Furthermore, we employ t-distributed stochastic neighbor embedding to visually explain the data distribution after classification. Experimental results demonstrate that cross en-tropy is a highly useful metric for evaluating the reliability of image classifier models. It is noted that for a more comprehensive evaluation of model perfor-mance, combining with other evaluation metrics is considered essential. .展开更多
This paper investigates the impact of reducing feature-vector dimensionality on the performance of machine learning(ML)models.Dimensionality reduction and feature selection techniques can improve computational efficie...This paper investigates the impact of reducing feature-vector dimensionality on the performance of machine learning(ML)models.Dimensionality reduction and feature selection techniques can improve computational efficiency,accuracy,robustness,transparency,and interpretability of ML models.In high-dimensional data,where features outnumber training instances,redundant or irrelevant features introduce noise,hindering model generalization and accuracy.This study explores the effects of dimensionality reduction methods on binary classifier performance using network traffic data for cybersecurity applications.The paper examines how dimensionality reduction techniques influence classifier operation and performance across diverse performancemetrics for seven ML models.Four dimensionality reduction methods are evaluated:principal component analysis(PCA),singular value decomposition(SVD),univariate feature selection(UFS)using chi-square statistics,and feature selection based on mutual information(MI).The results suggest that direct feature selection can be more effective than data projection methods in some applications.Direct selection offers lower computational complexity and,in some cases,superior classifier performance.This study emphasizes that evaluation and comparison of binary classifiers depend on specific performance metrics,each providing insights into different aspects of ML model operation.Using open-source network traffic data,this paper demonstrates that dimensionality reduction can be a valuable tool.It reduces computational overhead,enhances model interpretability and transparency,and maintains or even improves the performance of trained classifiers.The study also reveals that direct feature selection can be a more effective strategy when compared to feature engineering in specific scenarios.展开更多
Performance metrics and models are prerequisites for scientific understanding and optimization. This paper introduces a new footprint-based theory and reviews the research in the past four decades leading to the new t...Performance metrics and models are prerequisites for scientific understanding and optimization. This paper introduces a new footprint-based theory and reviews the research in the past four decades leading to the new theory. The review groups the past work into metrics and their models in particular those of the reuse distance, metrics conversion, models of shared cache, performance and optimization, and other related techniques.展开更多
The transportation and logistics sectors are major contributors to Greenhouse Gase(GHG)emissions.Carbon dioxide(CO_(2))from Light-Duty Vehicles(LDVs)is posing serious risks to air quality and public health.Understandi...The transportation and logistics sectors are major contributors to Greenhouse Gase(GHG)emissions.Carbon dioxide(CO_(2))from Light-Duty Vehicles(LDVs)is posing serious risks to air quality and public health.Understanding the extent of LDVs’impact on climate change and human well-being is crucial for informed decisionmaking and effective mitigation strategies.This study investigates the predictability of CO_(2)emissions from LDVs using a comprehensive dataset that includes vehicles from various manufacturers,their CO_(2)emission levels,and key influencing factors.Specifically,sixMachine Learning(ML)algorithms,ranging fromsimple linearmodels to complex non-linear models,were applied under identical conditions to ensure a fair comparison and their performance metrics were calculated.The obtained results showed a significant influence of variables such as engine size on CO_(2)emissions.Although the six algorithms have provided accurate forecasts,the Linear Regression(LR)model was found to be sufficient,achieving a Mean Absolute Percentage Error(MAPE)below 0.90%and a Coefficient of Determination(R2)exceeding 99.7%.These findings may contribute to a deeper understanding of LDVs’role in CO_(2)emissions and offer actionable insights for reducing their environmental impact.In fact,vehicle manufacturers can leverage these insights to target key emission-related factors,while policymakers and stakeholders in logistics and transportation can use the models to estimate the CO_(2)emissions of new vehicles before their market deployment or to project future emissions from current and expected LDV fleets.展开更多
We apply reweighted inversion focusing to full tensor gravity gradiometry data using message-passing interface （MPI） and compute unified device architecture （CUDA） parallel computing algorithms, and then combine M...We apply reweighted inversion focusing to full tensor gravity gradiometry data using message-passing interface （MPI） and compute unified device architecture （CUDA） parallel computing algorithms, and then combine MPI with CUDA to formulate a hybrid algorithm. Parallel computing performance metrics are introduced to analyze and compare the performance of the algorithms. We summarize the rules for the performance evaluation of parallel algorithms. We use model and real data from the Vinton salt dome to test the algorithms. We find good match between model efficiency and feasibility of parallel computing gravity gradiometry data. and real density data, and verify the high algorithms in the inversion of full tensor展开更多
Zinc–bromine rechargeable batteries(ZBRBs)are one of the most powerful candidates for next-generation energy storage due to their potentially lower material cost,deep discharge capability,non-flammable electrolytes,r...Zinc–bromine rechargeable batteries(ZBRBs)are one of the most powerful candidates for next-generation energy storage due to their potentially lower material cost,deep discharge capability,non-flammable electrolytes,relatively long lifetime and good reversibility.However,many opportunities remain to improve the efficiency and stability of these batteries for long-life operation.Here,we discuss the device configurations,working mechanisms and performance evaluation of ZBRBs.Both non-flow(static)and flow-type cells are highlighted in detail in this review.The fundamental electrochemical aspects,including the key challenges and promising solutions,are discussed,with particular attention paid to zinc and bromine half-cells,as their performance plays a critical role in determining the electrochemical performance of the battery system.The following sections examine the key performance metrics of ZBRBs and assessment methods using various ex situ and in situ/operando techniques.The review concludes with insights into future developments and prospects for high-performance ZBRBs.展开更多
In order to attain better communications performance rather than just expand coverage and save system cost,criteria related to the communications quality and capacity are extracted and revised to build an integrated p...In order to attain better communications performance rather than just expand coverage and save system cost,criteria related to the communications quality and capacity are extracted and revised to build an integrated performance metric system which aims to effectively guide the satellite communications constellation design.These performance metrics together with the system cost serve as the multiple objectives whilst the coverage requirement is regarded as the basic constraint in the optimization of the constellation configuration design applying a revised NSGA-II algorithm.The Pareto hyper-volumes lead to the best configuration schemes which achieve better integrated system performance compared with the conventional design results based merely on coverage and cost.展开更多
The purpose of this article is to receive environmental assessments of combustion of different types of coal fuel depending on the preparation(unscreened,size-graded,briquetted and heat-treated)in automated boilers an...The purpose of this article is to receive environmental assessments of combustion of different types of coal fuel depending on the preparation(unscreened,size-graded,briquetted and heat-treated)in automated boilers and boilers with manual load-ing.The assessments were made on the basis of data obtained from experimental methods of coal preparation and calculated methods of determining the amount of pollutant and greenhouse gas emissions,as well as the mass of ash and slag waste.The main pollutants from coal combustion are calculated:particulate matter,benz(a)pyrene,nitrogen oxides,sulfur dioxide,carbon monoxide.Of the greenhouse gases carbon dioxide is calculated.As a result of conducted research it is shown that the simplest preliminary preparation(size-graded)of coal significantly improves combustion efficiency and environmental performance:emissions are reduced by 13%for hard coal and up to 20%for brown coal.The introduction of automated boil-ers with heat-treated coal in small boiler facilities allows to reduce emissions and ash and slag waste by 2-3 times.The best environmental indicators correspond to heat-treated lignite,which is characterized by the absence of sulfur dioxide emissions.展开更多
Video streaming applications have grown considerably in recent years.As a result,this becomes one of the most significant contributors to global internet traffic.According to recent studies,the telecommunications indu...Video streaming applications have grown considerably in recent years.As a result,this becomes one of the most significant contributors to global internet traffic.According to recent studies,the telecommunications industry loses millions of dollars due to poor video Quality of Experience(QoE)for users.Among the standard proposals for standardizing the quality of video streaming over internet service providers(ISPs)is the Mean Opinion Score(MOS).However,the accurate finding of QoE by MOS is subjective and laborious,and it varies depending on the user.A fully automated data analytics framework is required to reduce the inter-operator variability characteristic in QoE assessment.This work addresses this concern by suggesting a novel hybrid XGBStackQoE analytical model using a two-level layering technique.Level one combines multiple Machine Learning(ML)models via a layer one Hybrid XGBStackQoE-model.Individual ML models at level one are trained using the entire training data set.The level two Hybrid XGBStackQoE-Model is fitted using the outputs(meta-features)of the layer one ML models.The proposed model outperformed the conventional models,with an accuracy improvement of 4 to 5 percent,which is still higher than the current traditional models.The proposed framework could significantly improve video QoE accuracy.展开更多
Forecasting river flow is crucial for optimal planning,management,and sustainability using freshwater resources.Many machine learning(ML)approaches have been enhanced to improve streamflow prediction.Hybrid techniques...Forecasting river flow is crucial for optimal planning,management,and sustainability using freshwater resources.Many machine learning(ML)approaches have been enhanced to improve streamflow prediction.Hybrid techniques have been viewed as a viable method for enhancing the accuracy of univariate streamflow estimation when compared to standalone approaches.Current researchers have also emphasised using hybrid models to improve forecast accuracy.Accordingly,this paper conducts an updated literature review of applications of hybrid models in estimating streamflow over the last five years,summarising data preprocessing,univariate machine learning modelling strategy,advantages and disadvantages of standalone ML techniques,hybrid models,and performance metrics.This study focuses on two types of hybrid models:parameter optimisation-based hybrid models(OBH)and hybridisation of parameter optimisation-based and preprocessing-based hybridmodels(HOPH).Overall,this research supports the idea thatmeta-heuristic approaches precisely improveML techniques.It’s also one of the first efforts to comprehensively examine the efficiency of various meta-heuristic approaches(classified into four primary classes)hybridised with ML techniques.This study revealed that previous research applied swarm,evolutionary,physics,and hybrid metaheuristics with 77%,61%,12%,and 12%,respectively.Finally,there is still room for improving OBH and HOPH models by examining different data pre-processing techniques and metaheuristic algorithms.展开更多
Traditional e-commerce recommendation systems often struggle with dynamic user preferences and a vast array of products,leading to suboptimal user experiences.To address this,our study presents a Personalized Adaptive...Traditional e-commerce recommendation systems often struggle with dynamic user preferences and a vast array of products,leading to suboptimal user experiences.To address this,our study presents a Personalized Adaptive Multi-Product Recommendation System(PAMR)leveraging transfer learning and Bi-GRU(Bidirectional Gated Recurrent Units).Using a large dataset of user reviews from Amazon and Flipkart,we employ transfer learning with pre-trained models(AlexNet,GoogleNet,ResNet-50)to extract high-level attributes from product data,ensuring effective feature representation even with limited data.Bi-GRU captures both spatial and sequential dependencies in user-item interactions.The innovation of this study lies in the innovative feature fusion technique that combines the strengths of multiple transfer learning models,and the integration of an attention mechanism within the Bi-GRU framework to prioritize relevant features.Our approach addresses the classic recommendation systems that often face challenges such as cold start along with data sparsity difficulties,by utilizing robust user and item representations.The model demonstrated an accuracy of up to 96.9%,with precision and an F1-score of 96.2%and 96.97%,respectively,on the Amazon dataset,significantly outperforming the baselines and marking a considerable advancement over traditional configurations.This study highlights the effectiveness of combining transfer learning with Bi-GRU for scalable and adaptive recommendation systems,providing a versatile solution for real-world applications.展开更多
This paper provides a detailed mathematical model governing the operation of feedforward neural networks(FFNN)and derives the backpropagation formulation utilized in the training process.Network protection systems mus...This paper provides a detailed mathematical model governing the operation of feedforward neural networks(FFNN)and derives the backpropagation formulation utilized in the training process.Network protection systems must ensure secure access to the Internet,reliability of network services,consistency of applications,safeguarding of stored information,and data integrity while in transit across networks.The paper reports on the application of neural networks(NN)and deep learning(DL)analytics to the detection of network traffic anomalies,including network intrusions,and the timely prevention and mitigation of cyberattacks.Among the most prevalent cyber threats are R2L,U2L,probe,and distributed denial of service(DDoS),which disrupt normal network operations and interrupt vital services.Robust detection of the early stage of cyberattack phenomena and the consistent blockade of attack traffic including DDoS network packets comprise preventive measures that constitute effective means for cyber defense.The proposed system is an NN that utilizes a set of thirty-eight packet features for the real-time binary classification of network traffic.The NN system is trained with a dataset containing the packet attributes of a mix of normal and attack traffic.In this study,the KDD99 dataset,which was prepared by the MIT Lincoln Lab for the 1998 DARPA Intrusion Detection Evaluation Program,was used to train the NN and test its performance.It has been shown that an NN comprised of one or two hidden layers,with each layer containing a few neural nodes,can be trained to detect attack packets with concurrently high precision and recall.展开更多
This paper presents a novel general method for computing optimal motions of an industrial robot manipulator （AdeptOne XL robot） in the presence of fixed and oscillating obstacles. The optimization model considers th...This paper presents a novel general method for computing optimal motions of an industrial robot manipulator （AdeptOne XL robot） in the presence of fixed and oscillating obstacles. The optimization model considers the nonlinear manipulator dynamics, actuator constraints, joint limits, and obstacle avoidance. The problem has 6 objective functions, 88 variables, and 21 constraints. Two evolutionary algorithms, namely, elitist non-dominated sorting genetic algorithm （NSGA-II） and multi-objective differential evolution （MODE）, have been used for the optimization. Two methods （normalized weighting objective functions and average fitness factor） are used to select the best solution tradeoffs. Two multi-objective performance measures, namely solution spread measure and ratio of non-dominated individuals, are used to evaluate the Pareto optimal fronts. Two multi-objective performance measures, namely, optimizer overhead and algorithm effort, are used to find the computational effort of the optimization algorithm. The trajectories are defined by B-spline functions. The results obtained from NSGA-II and MODE are compared and analyzed.展开更多
Fog computing brings computational services near the network edge to meet the latency constraints of cyber-physical System(CPS)applications.Edge devices enable limited computational capacity and energy availability th...Fog computing brings computational services near the network edge to meet the latency constraints of cyber-physical System(CPS)applications.Edge devices enable limited computational capacity and energy availability that hamper end user performance.We designed a novel performance measurement index to gauge a device’s resource capacity.This examination addresses the offloading mechanism issues,where the end user(EU)offloads a part of its workload to a nearby edge server(ES).Sometimes,the ES further offloads the workload to another ES or cloud server to achieve reliable performance because of limited resources(such as storage and computation).The manuscript aims to reduce the service offloading rate by selecting a potential device or server to accomplish a low average latency and service completion time to meet the deadline constraints of sub-divided services.In this regard,an adaptive online status predictive model design is significant for prognosticating the asset requirement of arrived services to make float decisions.Consequently,the development of a reinforcement learning-based flexible x-scheduling(RFXS)approach resolves the service offloading issues,where x=service/resource for producing the low latency and high performance of the network.Our approach to the theoretical bound and computational complexity is derived by formulating the system efficiency.A quadratic restraint mechanism is employed to formulate the service optimization issue according to a set ofmeasurements,as well as the behavioural association rate and adulation factor.Our system managed an average 0.89%of the service offloading rate,with 39 ms of delay over complex scenarios(using three servers with a 50%service arrival rate).The simulation outcomes confirm that the proposed scheme attained a low offloading uncertainty,and is suitable for simulating heterogeneous CPS frameworks.展开更多
This paper proposes a cryptographic technique on images based on the Sudoku solution.Sudoku is a number puzzle,which needs applying defined protocols and filling the empty boxes with numbers.Given a small size of numb...This paper proposes a cryptographic technique on images based on the Sudoku solution.Sudoku is a number puzzle,which needs applying defined protocols and filling the empty boxes with numbers.Given a small size of numbers as input,solving the sudoku puzzle yields an expanded big size of numbers,which can be used as a key for the Encryption/Decryption of images.In this way,the given small size of numbers can be stored as the prime key,which means the key is compact.A prime key clue in the sudoku puzzle always leads to only one solution,which means the key is always stable.This feature is the background for the paper,where the Sudoku puzzle output can be innovatively introduced in image cryptography.Sudoku solution is expanded to any size image using a sequence of expansion techniques that involve filling of the number matrix,Linear X-Y rotational shifting,and reverse shifting based on a standard zig-zag pattern.The crypto key for an image dictates the details of positions,where the image pixels have to be shuffled.Shuffling is made at two levels,namely pixel and sub-pixel(RGB)levels for an image,with the latter having more effective Encryption.The brought-out technique falls under the Image scrambling method with partial diffusion.Performance metrics are impressive and are given by a Histogram deviation of 0.997,a Correlation coefficient of 10−2 and an NPCR of 99.98%.Hence,it is evident that the image cryptography with the sudoku kept in place is more efficient against Plaintext and Differential attacks.展开更多
The rapid growth of interconnected high performance workstations has produced a new computing paradigm called clustered of workstations computing. In these systems load balance problem is a serious impediment to achie...The rapid growth of interconnected high performance workstations has produced a new computing paradigm called clustered of workstations computing. In these systems load balance problem is a serious impediment to achieve good performance. The main concern of this paper is the implementation of dynamic load balancing algorithm, asynchronous Round Robin (ARR), for balancing workload of parallel tree computation depth-first-search algorithm on Cluster of Heterogeneous Workstations (COW) Many algorithms in artificial intelligence and other areas of computer science are based on depth first search in implicitty defined trees. For these algorithms a load-balancing scheme is required, which is able to evenly distribute parts of an irregularly shaped tree over the workstations with minimal interprocessor communication and without prior knowledge of the tree’s shape. For the (ARR) algorithm only minimal interprocessor communication is needed when necessary and it runs under the MPI (Message passing interface) that allows parallel execution on heterogeneous SUN cluster of workstation platform. The program code is written in C language and executed under UNIX operating system (Solaris version).展开更多
Cryptocurrency price prediction has garnered significant attention due to the growing importance of digital assets in the financial landscape. This paper presents a comprehensive study on predicting future cryptocurre...Cryptocurrency price prediction has garnered significant attention due to the growing importance of digital assets in the financial landscape. This paper presents a comprehensive study on predicting future cryptocurrency prices using machine learning algorithms. Open-source historical data from various cryptocurrency exchanges is utilized. Interpolation techniques are employed to handle missing data, ensuring the completeness and reliability of the dataset. Four technical indicators are selected as features for prediction. The study explores the application of five machine learning algorithms to capture the complex patterns in the highly volatile cryptocurrency market. The findings demonstrate the strengths and limitations of the different approaches, highlighting the significance of feature engineering and algorithm selection in achieving accurate cryptocurrency price predictions. The research contributes valuable insights into the dynamic and rapidly evolving field of cryptocurrency price prediction, assisting investors and traders in making informed decisions amidst the challenges posed by the cryptocurrency market.展开更多
Existing systems use key performance indicators(KPIs)as metrics for physical layer(PHY)optimization,which suffers from the problem of overoptimization,because some unnecessary PHY enhancements are imperceptible to ter...Existing systems use key performance indicators(KPIs)as metrics for physical layer(PHY)optimization,which suffers from the problem of overoptimization,because some unnecessary PHY enhancements are imperceptible to terminal users and thus induce additional cost and energy waste.Therefore,it is necessary to utilize directly the quality of experience(QoE)of user as a metric of optimization,which can achieve the global optimum of QoE under cost and energy constraints.However,QoE is still a metric of application layer that cannot be easily used to design and optimize the PHY.To address this problem,we in this paper propose a novel end-to-end QoE(E2E-QoE)based optimization architecture at the user-side for the first time.Specifically,a cross-layer parameterized model is proposed to establish the relationship between PHY and E2E-QoE.Based on this,an E2E-QoE oriented PHY anomaly diagnosis method is further designed to locate the time and root cause of anomalies.Finally,we investigate to optimize the PHY algorithm directly based on the E2E-QoE.The proposed frameworks and algorithms are all validated using the data from real fifth-generation(5G)mobile system,which show that using E2E-QoE as the metric of PHY optimization is feasible and can outperform existing schemes.展开更多
Greenroads(www.greenroads.us)is a performance metric for sustainable practices associated with the design and construction of roads.It assigns points for approved sustainable choices/practices and can be used to asses...Greenroads(www.greenroads.us)is a performance metric for sustainable practices associated with the design and construction of roads.It assigns points for approved sustainable choices/practices and can be used to assess roadway project sustainability measures based on total points.Such a metric can(1)provide a quantitative means of sustainability assessment,(2)allow informed sustainability decisions,(3)provide baseline sustainability standards,and(4)stimulate improvement and innovation in integrated roadway sustainability.This paper describes Greenroads version 1.0,which consists of 11 requirements and 37 voluntary practices that can be used as a project-level sustainability performance metric.Development efforts and a Washington State Department of Transportation(WSDOT)case study suggest(1)existing project data can serve as the data source for performance assessment,(2)some requirements and voluntary actions need refinement,(3)projects need to treat sustainability in a holistic manner to meet a reasonable sustainability performance standard,(4),the financial impact of Greenroads use must be studied,and(5)several pilot projects are needed.The Greenroads sustainability performance metric can be a viable means of projectlevel sustainability performance assessment and decision support.展开更多
文摘The new measurement scheme of IP performance metrics is for the mobile network in heterogeneous wireless network environment. In the proposed scheme, when Mobile Nodes （MNs） inside the mobile network needs to understand the condition of multiple comrmunicatinn paths outside the mobile netwtrk, they can get IP performance metrics, such as delay, jitter, bandwidth, packet loss, etc., irrespective of the preserre or absence of measurement functionality. At the same time, the proposed scheme dees not require the MN to he involved in measuring IP performance metrice. The Multihomed Mobile Router （MMR） with heterogeneons wireless interfaces measures IP performance metrics on behalf of the MNs inside the mobile network. Then, MNs can get measured IP perfonmnce metries from the MMR using L3 messages. The proposed scheme can reduce burden and power consumption of MNs with limited resource and batty power since MNs don＇ t measure IP performance metrics directly. In addition, it can reduce considerably traffic overhead over wireless links on multiple measurement paths since signaling messages and injeeted testing traffic are reduced.
文摘Cross entropy is a measure in machine learning and deep learning that assesses the difference between predicted and actual probability distributions. In this study, we propose cross entropy as a performance evaluation metric for image classifier models and apply it to the CT image classification of lung cancer. A convolutional neural network is employed as the deep neural network (DNN) image classifier, with the residual network (ResNet) 50 chosen as the DNN archi-tecture. The image data used comprise a lung CT image set. Two classification models are built from datasets with varying amounts of data, and lung cancer is categorized into four classes using 10-fold cross-validation. Furthermore, we employ t-distributed stochastic neighbor embedding to visually explain the data distribution after classification. Experimental results demonstrate that cross en-tropy is a highly useful metric for evaluating the reliability of image classifier models. It is noted that for a more comprehensive evaluation of model perfor-mance, combining with other evaluation metrics is considered essential. .
基金funded by US Army Combat Capabilities Development Command(CCDC)Aviation&Missile Center,https://www.avmc.army.mil/(accessed on 5 February 2024),CONTRACT NUMBER:W31P4Q-18-D-0002 through Georgia Tech Research Institute and AAMU-RISE。
文摘This paper investigates the impact of reducing feature-vector dimensionality on the performance of machine learning(ML)models.Dimensionality reduction and feature selection techniques can improve computational efficiency,accuracy,robustness,transparency,and interpretability of ML models.In high-dimensional data,where features outnumber training instances,redundant or irrelevant features introduce noise,hindering model generalization and accuracy.This study explores the effects of dimensionality reduction methods on binary classifier performance using network traffic data for cybersecurity applications.The paper examines how dimensionality reduction techniques influence classifier operation and performance across diverse performancemetrics for seven ML models.Four dimensionality reduction methods are evaluated:principal component analysis(PCA),singular value decomposition(SVD),univariate feature selection(UFS)using chi-square statistics,and feature selection based on mutual information(MI).The results suggest that direct feature selection can be more effective than data projection methods in some applications.Direct selection offers lower computational complexity and,in some cases,superior classifier performance.This study emphasizes that evaluation and comparison of binary classifiers depend on specific performance metrics,each providing insights into different aspects of ML model operation.Using open-source network traffic data,this paper demonstrates that dimensionality reduction can be a valuable tool.It reduces computational overhead,enhances model interpretability and transparency,and maintains or even improves the performance of trained classifiers.The study also reveals that direct feature selection can be a more effective strategy when compared to feature engineering in specific scenarios.
基金partially supported by the National Natural Science Foundation of China(NSFC)under Grant No.61232008the NSFC Joint Research Fund for Overseas Chinese Scholars and Scholars in Hong Kong and Macao under Grant No.61328201+2 种基金the National Science Foundation of USA under Contract Nos.CNS-1319617,CCF-1116104,CCF-0963759an IBM CAS Faculty Fellowshipa research grant from Huawei
文摘Performance metrics and models are prerequisites for scientific understanding and optimization. This paper introduces a new footprint-based theory and reviews the research in the past four decades leading to the new theory. The review groups the past work into metrics and their models in particular those of the reuse distance, metrics conversion, models of shared cache, performance and optimization, and other related techniques.
基金Deputyship for Research&Innovation,Ministry of Education in Saudi Arabia,project number MoE-IF-UJ-R2-22-20772-1.
文摘The transportation and logistics sectors are major contributors to Greenhouse Gase(GHG)emissions.Carbon dioxide(CO_(2))from Light-Duty Vehicles(LDVs)is posing serious risks to air quality and public health.Understanding the extent of LDVs’impact on climate change and human well-being is crucial for informed decisionmaking and effective mitigation strategies.This study investigates the predictability of CO_(2)emissions from LDVs using a comprehensive dataset that includes vehicles from various manufacturers,their CO_(2)emission levels,and key influencing factors.Specifically,sixMachine Learning(ML)algorithms,ranging fromsimple linearmodels to complex non-linear models,were applied under identical conditions to ensure a fair comparison and their performance metrics were calculated.The obtained results showed a significant influence of variables such as engine size on CO_(2)emissions.Although the six algorithms have provided accurate forecasts,the Linear Regression(LR)model was found to be sufficient,achieving a Mean Absolute Percentage Error(MAPE)below 0.90%and a Coefficient of Determination(R2)exceeding 99.7%.These findings may contribute to a deeper understanding of LDVs’role in CO_(2)emissions and offer actionable insights for reducing their environmental impact.In fact,vehicle manufacturers can leverage these insights to target key emission-related factors,while policymakers and stakeholders in logistics and transportation can use the models to estimate the CO_(2)emissions of new vehicles before their market deployment or to project future emissions from current and expected LDV fleets.
基金supported by the Sino-Probe09(No.201011078)National High-tech R&D Program(No.863 and2014AA06A613)
文摘We apply reweighted inversion focusing to full tensor gravity gradiometry data using message-passing interface （MPI） and compute unified device architecture （CUDA） parallel computing algorithms, and then combine MPI with CUDA to formulate a hybrid algorithm. Parallel computing performance metrics are introduced to analyze and compare the performance of the algorithms. We summarize the rules for the performance evaluation of parallel algorithms. We use model and real data from the Vinton salt dome to test the algorithms. We find good match between model efficiency and feasibility of parallel computing gravity gradiometry data. and real density data, and verify the high algorithms in the inversion of full tensor
基金flnancial support from Australian Research Council through its Discovery,Future Fellowship ProgramsImam Mohammad Ibn Saud Islamic University (IMSIU) in Riyadh,Saudi Arabia,for flnancial support of this work.
文摘Zinc–bromine rechargeable batteries(ZBRBs)are one of the most powerful candidates for next-generation energy storage due to their potentially lower material cost,deep discharge capability,non-flammable electrolytes,relatively long lifetime and good reversibility.However,many opportunities remain to improve the efficiency and stability of these batteries for long-life operation.Here,we discuss the device configurations,working mechanisms and performance evaluation of ZBRBs.Both non-flow(static)and flow-type cells are highlighted in detail in this review.The fundamental electrochemical aspects,including the key challenges and promising solutions,are discussed,with particular attention paid to zinc and bromine half-cells,as their performance plays a critical role in determining the electrochemical performance of the battery system.The following sections examine the key performance metrics of ZBRBs and assessment methods using various ex situ and in situ/operando techniques.The review concludes with insights into future developments and prospects for high-performance ZBRBs.
文摘In order to attain better communications performance rather than just expand coverage and save system cost,criteria related to the communications quality and capacity are extracted and revised to build an integrated performance metric system which aims to effectively guide the satellite communications constellation design.These performance metrics together with the system cost serve as the multiple objectives whilst the coverage requirement is regarded as the basic constraint in the optimization of the constellation configuration design applying a revised NSGA-II algorithm.The Pareto hyper-volumes lead to the best configuration schemes which achieve better integrated system performance compared with the conventional design results based merely on coverage and cost.
基金The research was carried out under State Assignment Projects(FWEU-2021-0004,FWEU-2021-0005)of the Fundamental Research Program of Russian Federation 2021-2030.
文摘The purpose of this article is to receive environmental assessments of combustion of different types of coal fuel depending on the preparation(unscreened,size-graded,briquetted and heat-treated)in automated boilers and boilers with manual load-ing.The assessments were made on the basis of data obtained from experimental methods of coal preparation and calculated methods of determining the amount of pollutant and greenhouse gas emissions,as well as the mass of ash and slag waste.The main pollutants from coal combustion are calculated:particulate matter,benz(a)pyrene,nitrogen oxides,sulfur dioxide,carbon monoxide.Of the greenhouse gases carbon dioxide is calculated.As a result of conducted research it is shown that the simplest preliminary preparation(size-graded)of coal significantly improves combustion efficiency and environmental performance:emissions are reduced by 13%for hard coal and up to 20%for brown coal.The introduction of automated boil-ers with heat-treated coal in small boiler facilities allows to reduce emissions and ash and slag waste by 2-3 times.The best environmental indicators correspond to heat-treated lignite,which is characterized by the absence of sulfur dioxide emissions.
文摘Video streaming applications have grown considerably in recent years.As a result,this becomes one of the most significant contributors to global internet traffic.According to recent studies,the telecommunications industry loses millions of dollars due to poor video Quality of Experience(QoE)for users.Among the standard proposals for standardizing the quality of video streaming over internet service providers(ISPs)is the Mean Opinion Score(MOS).However,the accurate finding of QoE by MOS is subjective and laborious,and it varies depending on the user.A fully automated data analytics framework is required to reduce the inter-operator variability characteristic in QoE assessment.This work addresses this concern by suggesting a novel hybrid XGBStackQoE analytical model using a two-level layering technique.Level one combines multiple Machine Learning(ML)models via a layer one Hybrid XGBStackQoE-model.Individual ML models at level one are trained using the entire training data set.The level two Hybrid XGBStackQoE-Model is fitted using the outputs(meta-features)of the layer one ML models.The proposed model outperformed the conventional models,with an accuracy improvement of 4 to 5 percent,which is still higher than the current traditional models.The proposed framework could significantly improve video QoE accuracy.
基金This paper’s logical organisation and content quality have been enhanced,so the authors thank anonymous reviewers and journal editors for assistance.
文摘Forecasting river flow is crucial for optimal planning,management,and sustainability using freshwater resources.Many machine learning(ML)approaches have been enhanced to improve streamflow prediction.Hybrid techniques have been viewed as a viable method for enhancing the accuracy of univariate streamflow estimation when compared to standalone approaches.Current researchers have also emphasised using hybrid models to improve forecast accuracy.Accordingly,this paper conducts an updated literature review of applications of hybrid models in estimating streamflow over the last five years,summarising data preprocessing,univariate machine learning modelling strategy,advantages and disadvantages of standalone ML techniques,hybrid models,and performance metrics.This study focuses on two types of hybrid models:parameter optimisation-based hybrid models(OBH)and hybridisation of parameter optimisation-based and preprocessing-based hybridmodels(HOPH).Overall,this research supports the idea thatmeta-heuristic approaches precisely improveML techniques.It’s also one of the first efforts to comprehensively examine the efficiency of various meta-heuristic approaches(classified into four primary classes)hybridised with ML techniques.This study revealed that previous research applied swarm,evolutionary,physics,and hybrid metaheuristics with 77%,61%,12%,and 12%,respectively.Finally,there is still room for improving OBH and HOPH models by examining different data pre-processing techniques and metaheuristic algorithms.
文摘Traditional e-commerce recommendation systems often struggle with dynamic user preferences and a vast array of products,leading to suboptimal user experiences.To address this,our study presents a Personalized Adaptive Multi-Product Recommendation System(PAMR)leveraging transfer learning and Bi-GRU(Bidirectional Gated Recurrent Units).Using a large dataset of user reviews from Amazon and Flipkart,we employ transfer learning with pre-trained models(AlexNet,GoogleNet,ResNet-50)to extract high-level attributes from product data,ensuring effective feature representation even with limited data.Bi-GRU captures both spatial and sequential dependencies in user-item interactions.The innovation of this study lies in the innovative feature fusion technique that combines the strengths of multiple transfer learning models,and the integration of an attention mechanism within the Bi-GRU framework to prioritize relevant features.Our approach addresses the classic recommendation systems that often face challenges such as cold start along with data sparsity difficulties,by utilizing robust user and item representations.The model demonstrated an accuracy of up to 96.9%,with precision and an F1-score of 96.2%and 96.97%,respectively,on the Amazon dataset,significantly outperforming the baselines and marking a considerable advancement over traditional configurations.This study highlights the effectiveness of combining transfer learning with Bi-GRU for scalable and adaptive recommendation systems,providing a versatile solution for real-world applications.
基金Kaveh Heidary’s research project was partially funded by Quantum ResearchInternational Inc.through Contract QRI-SC-20-105.https://www.quantum-intl.com/.
文摘This paper provides a detailed mathematical model governing the operation of feedforward neural networks(FFNN)and derives the backpropagation formulation utilized in the training process.Network protection systems must ensure secure access to the Internet,reliability of network services,consistency of applications,safeguarding of stored information,and data integrity while in transit across networks.The paper reports on the application of neural networks(NN)and deep learning(DL)analytics to the detection of network traffic anomalies,including network intrusions,and the timely prevention and mitigation of cyberattacks.Among the most prevalent cyber threats are R2L,U2L,probe,and distributed denial of service(DDoS),which disrupt normal network operations and interrupt vital services.Robust detection of the early stage of cyberattack phenomena and the consistent blockade of attack traffic including DDoS network packets comprise preventive measures that constitute effective means for cyber defense.The proposed system is an NN that utilizes a set of thirty-eight packet features for the real-time binary classification of network traffic.The NN system is trained with a dataset containing the packet attributes of a mix of normal and attack traffic.In this study,the KDD99 dataset,which was prepared by the MIT Lincoln Lab for the 1998 DARPA Intrusion Detection Evaluation Program,was used to train the NN and test its performance.It has been shown that an NN comprised of one or two hidden layers,with each layer containing a few neural nodes,can be trained to detect attack packets with concurrently high precision and recall.
文摘This paper presents a novel general method for computing optimal motions of an industrial robot manipulator （AdeptOne XL robot） in the presence of fixed and oscillating obstacles. The optimization model considers the nonlinear manipulator dynamics, actuator constraints, joint limits, and obstacle avoidance. The problem has 6 objective functions, 88 variables, and 21 constraints. Two evolutionary algorithms, namely, elitist non-dominated sorting genetic algorithm （NSGA-II） and multi-objective differential evolution （MODE）, have been used for the optimization. Two methods （normalized weighting objective functions and average fitness factor） are used to select the best solution tradeoffs. Two multi-objective performance measures, namely solution spread measure and ratio of non-dominated individuals, are used to evaluate the Pareto optimal fronts. Two multi-objective performance measures, namely, optimizer overhead and algorithm effort, are used to find the computational effort of the optimization algorithm. The trajectories are defined by B-spline functions. The results obtained from NSGA-II and MODE are compared and analyzed.
基金Zulqar and Kim’s research was supported in part by the Basic Science Research Program through the National Research Foundation of Korea(NRF)funded by the Ministry of Education(NRF-2021R1A6A1A03039493)in part by the NRF grant funded by the Korea government(MSIT)(NRF-2022R1A2C1004401)+1 种基金Mekala’s research was supported in part by the Basic Science Research Program of the Ministry of Education(NRF-2018R1A2B6005105)in part by the National Research Foundation of Korea(NRF)grant funded by the Korea Government(MSIT)(no.2019R1A5A8080290).
文摘Fog computing brings computational services near the network edge to meet the latency constraints of cyber-physical System(CPS)applications.Edge devices enable limited computational capacity and energy availability that hamper end user performance.We designed a novel performance measurement index to gauge a device’s resource capacity.This examination addresses the offloading mechanism issues,where the end user(EU)offloads a part of its workload to a nearby edge server(ES).Sometimes,the ES further offloads the workload to another ES or cloud server to achieve reliable performance because of limited resources(such as storage and computation).The manuscript aims to reduce the service offloading rate by selecting a potential device or server to accomplish a low average latency and service completion time to meet the deadline constraints of sub-divided services.In this regard,an adaptive online status predictive model design is significant for prognosticating the asset requirement of arrived services to make float decisions.Consequently,the development of a reinforcement learning-based flexible x-scheduling(RFXS)approach resolves the service offloading issues,where x=service/resource for producing the low latency and high performance of the network.Our approach to the theoretical bound and computational complexity is derived by formulating the system efficiency.A quadratic restraint mechanism is employed to formulate the service optimization issue according to a set ofmeasurements,as well as the behavioural association rate and adulation factor.Our system managed an average 0.89%of the service offloading rate,with 39 ms of delay over complex scenarios(using three servers with a 50%service arrival rate).The simulation outcomes confirm that the proposed scheme attained a low offloading uncertainty,and is suitable for simulating heterogeneous CPS frameworks.
基金supported by the government of the Basque Country for the ELKARTEK21/10 KK-2021/00014 and ELKARTEK22/85 Research Programs,respectively。
文摘This paper proposes a cryptographic technique on images based on the Sudoku solution.Sudoku is a number puzzle,which needs applying defined protocols and filling the empty boxes with numbers.Given a small size of numbers as input,solving the sudoku puzzle yields an expanded big size of numbers,which can be used as a key for the Encryption/Decryption of images.In this way,the given small size of numbers can be stored as the prime key,which means the key is compact.A prime key clue in the sudoku puzzle always leads to only one solution,which means the key is always stable.This feature is the background for the paper,where the Sudoku puzzle output can be innovatively introduced in image cryptography.Sudoku solution is expanded to any size image using a sequence of expansion techniques that involve filling of the number matrix,Linear X-Y rotational shifting,and reverse shifting based on a standard zig-zag pattern.The crypto key for an image dictates the details of positions,where the image pixels have to be shuffled.Shuffling is made at two levels,namely pixel and sub-pixel(RGB)levels for an image,with the latter having more effective Encryption.The brought-out technique falls under the Image scrambling method with partial diffusion.Performance metrics are impressive and are given by a Histogram deviation of 0.997,a Correlation coefficient of 10−2 and an NPCR of 99.98%.Hence,it is evident that the image cryptography with the sudoku kept in place is more efficient against Plaintext and Differential attacks.
文摘The rapid growth of interconnected high performance workstations has produced a new computing paradigm called clustered of workstations computing. In these systems load balance problem is a serious impediment to achieve good performance. The main concern of this paper is the implementation of dynamic load balancing algorithm, asynchronous Round Robin (ARR), for balancing workload of parallel tree computation depth-first-search algorithm on Cluster of Heterogeneous Workstations (COW) Many algorithms in artificial intelligence and other areas of computer science are based on depth first search in implicitty defined trees. For these algorithms a load-balancing scheme is required, which is able to evenly distribute parts of an irregularly shaped tree over the workstations with minimal interprocessor communication and without prior knowledge of the tree’s shape. For the (ARR) algorithm only minimal interprocessor communication is needed when necessary and it runs under the MPI (Message passing interface) that allows parallel execution on heterogeneous SUN cluster of workstation platform. The program code is written in C language and executed under UNIX operating system (Solaris version).
文摘Cryptocurrency price prediction has garnered significant attention due to the growing importance of digital assets in the financial landscape. This paper presents a comprehensive study on predicting future cryptocurrency prices using machine learning algorithms. Open-source historical data from various cryptocurrency exchanges is utilized. Interpolation techniques are employed to handle missing data, ensuring the completeness and reliability of the dataset. Four technical indicators are selected as features for prediction. The study explores the application of five machine learning algorithms to capture the complex patterns in the highly volatile cryptocurrency market. The findings demonstrate the strengths and limitations of the different approaches, highlighting the significance of feature engineering and algorithm selection in achieving accurate cryptocurrency price predictions. The research contributes valuable insights into the dynamic and rapidly evolving field of cryptocurrency price prediction, assisting investors and traders in making informed decisions amidst the challenges posed by the cryptocurrency market.
文摘Existing systems use key performance indicators(KPIs)as metrics for physical layer(PHY)optimization,which suffers from the problem of overoptimization,because some unnecessary PHY enhancements are imperceptible to terminal users and thus induce additional cost and energy waste.Therefore,it is necessary to utilize directly the quality of experience(QoE)of user as a metric of optimization,which can achieve the global optimum of QoE under cost and energy constraints.However,QoE is still a metric of application layer that cannot be easily used to design and optimize the PHY.To address this problem,we in this paper propose a novel end-to-end QoE(E2E-QoE)based optimization architecture at the user-side for the first time.Specifically,a cross-layer parameterized model is proposed to establish the relationship between PHY and E2E-QoE.Based on this,an E2E-QoE oriented PHY anomaly diagnosis method is further designed to locate the time and root cause of anomalies.Finally,we investigate to optimize the PHY algorithm directly based on the E2E-QoE.The proposed frameworks and algorithms are all validated using the data from real fifth-generation(5G)mobile system,which show that using E2E-QoE as the metric of PHY optimization is feasible and can outperform existing schemes.
文摘Greenroads(www.greenroads.us)is a performance metric for sustainable practices associated with the design and construction of roads.It assigns points for approved sustainable choices/practices and can be used to assess roadway project sustainability measures based on total points.Such a metric can(1)provide a quantitative means of sustainability assessment,(2)allow informed sustainability decisions,(3)provide baseline sustainability standards,and(4)stimulate improvement and innovation in integrated roadway sustainability.This paper describes Greenroads version 1.0,which consists of 11 requirements and 37 voluntary practices that can be used as a project-level sustainability performance metric.Development efforts and a Washington State Department of Transportation(WSDOT)case study suggest(1)existing project data can serve as the data source for performance assessment,(2)some requirements and voluntary actions need refinement,(3)projects need to treat sustainability in a holistic manner to meet a reasonable sustainability performance standard,(4),the financial impact of Greenroads use must be studied,and(5)several pilot projects are needed.The Greenroads sustainability performance metric can be a viable means of projectlevel sustainability performance assessment and decision support.
