A typical Whipple shield consists of double-layered plates with a certain gap.The space debris impacts the outer plate and is broken into a debris cloud(shattered,molten,vaporized)with dispersed energy and momentum,wh...A typical Whipple shield consists of double-layered plates with a certain gap.The space debris impacts the outer plate and is broken into a debris cloud(shattered,molten,vaporized)with dispersed energy and momentum,which reduces the risk of penetrating the bulkhead.In the realm of hypervelocity impact,strain rate(>10^(5)s^(-1))effects are negligible,and fluid dynamics is employed to describe the impact process.Efficient numerical tools for precisely predicting the damage degree can greatly accelerate the design and optimization of advanced protective structures.Current hypervelocity impact research primarily focuses on the interaction between projectile and front plate and the movement of debris cloud.However,the damage mechanism of debris cloud impacts on rear plates-the critical threat component-remains underexplored owing to complex multi-physics processes and prohibitive computational costs.Existing approaches,ranging from semi-empirical equations to a machine learningbased ballistic limit prediction method,are constrained to binary penetration classification.Alternatively,the uneven data from experiments and simulations caused these methods to be ineffective when the projectile has irregular shapes and complicate flight attitude.Therefore,it is urgent to develop a new damage prediction method for predicting the rear plate damage,which can help to gain a deeper understanding of the damage mechanism.In this study,a machine learning(ML)method is developed to predict the damage distribution in the rear plate.Based on the unit velocity space,the discretized information of debris cloud and rear plate damage from rare simulation cases is used as input data for training the ML models,while the generalization ability for damage distribution prediction is tested by other simulation cases with different attack angles.The results demonstrate that the training and prediction accuracies using the Random Forest(RF)algorithm significantly surpass those using Artificial Neural Networks(ANNs)and Support Vector Machine(SVM).The RF-based model effectively identifies damage features in sparsely distributed debris cloud and cumulative effect.This study establishes an expandable new dataset that accommodates additional parameters to improve the prediction accuracy.Results demonstrate the model's ability to overcome data imbalance limitations through debris cloud features,enabling rapid and accurate rear plate damage prediction across wider scenarios with minimal data requirements.展开更多
Many digital platforms have employed free-content promotion strategies to deal with the high uncertainty levels regarding digital content products.However,the diversity of digital content products and user heterogenei...Many digital platforms have employed free-content promotion strategies to deal with the high uncertainty levels regarding digital content products.However,the diversity of digital content products and user heterogeneity in content preference may blur the impact of platform promotions across users and products.Therefore,free-content promotion strategies should be adapted to allocate marketing resources optimally and increase revenue.This study develops personal-ized free-content promotion strategies based on individual-level heterogeneous treatment effects and explores the causes of their heterogeneity,focusing on the moderating effect of user engagement-related variables.To this end,we utilize ran-dom field experimental data provided by a top Chinese e-book platform.We employ a framework that combines machine learning with econometric causal inference methods to estimate individual treatment effects and analyze their potential mechanisms.The analysis shows that,on average,free-content promotions lead to a significant increase in consumer pay-ments.However,the higher the level of user engagement,the lower the payment lift caused by promotions,as more-engaged users are more strongly affected by the cannibalization effect of free-content promotion.This study introduces a novel causal research design to help platforms improve their marketing strategies.展开更多
Deep neural networks(DNNs)have achieved great success in many data processing applications.However,high computational complexity and storage cost make deep learning difficult to be used on resource-constrained devices...Deep neural networks(DNNs)have achieved great success in many data processing applications.However,high computational complexity and storage cost make deep learning difficult to be used on resource-constrained devices,and it is not environmental-friendly with much power cost.In this paper,we focus on low-rank optimization for efficient deep learning techniques.In the space domain,DNNs are compressed by low rank approximation of the network parameters,which directly reduces the storage requirement with a smaller number of network parameters.In the time domain,the network parameters can be trained in a few subspaces,which enables efficient training for fast convergence.The model compression in the spatial domain is summarized into three categories as pre-train,pre-set,and compression-aware methods,respectively.With a series of integrable techniques discussed,such as sparse pruning,quantization,and entropy coding,we can ensemble them in an integration framework with lower computational complexity and storage.In addition to summary of recent technical advances,we have two findings for motivating future works.One is that the effective rank,derived from the Shannon entropy of the normalized singular values,outperforms other conventional sparse measures such as the?_1 norm for network compression.The other is a spatial and temporal balance for tensorized neural networks.For accelerating the training of tensorized neural networks,it is crucial to leverage redundancy for both model compression and subspace training.展开更多
As one of the most important part of weapon system of systems(WSoS),quantitative evaluation of reconnaissance satellite system(RSS)is indispensable during its construction and application.Aiming at the problem of nonl...As one of the most important part of weapon system of systems(WSoS),quantitative evaluation of reconnaissance satellite system(RSS)is indispensable during its construction and application.Aiming at the problem of nonlinear effectiveness evaluation under small sample conditions,we propose an evaluation method based on support vector regression(SVR)to effectively address the defects of traditional methods.Considering the performance of SVR is influenced by the penalty factor,kernel type,and other parameters deeply,the improved grey wolf optimizer(IGWO)is employed for parameter optimization.In the proposed IGWO algorithm,the opposition-based learning strategy is adopted to increase the probability of avoiding the local optima,the mutation operator is used to escape from premature convergence and differential convergence factors are applied to increase the rate of convergence.Numerical experiments of 14 test functions validate the applicability of IGWO algorithm dealing with global optimization.The index system and evaluation method are constructed based on the characteristics of RSS.To validate the proposed IGWO-SVR evaluation method,eight benchmark data sets and combat simulation are employed to estimate the evaluation accuracy,convergence performance and computational complexity.According to the experimental results,the proposed method outperforms several prediction based evaluation methods,verifies the superiority and effectiveness in RSS operational effectiveness evaluation.展开更多
日益频繁的鸟类活动给输电线路的安全运行带来了极大威胁,而现有拟声驱鸟装置由于缺乏智能性,无法长期有效驱鸟.为了解决上述问题,本文提出基于改进Q⁃learning算法的拟声驱鸟策略.首先,为了评价各音频的驱鸟效果,结合模糊理论,将鸟类听...日益频繁的鸟类活动给输电线路的安全运行带来了极大威胁,而现有拟声驱鸟装置由于缺乏智能性,无法长期有效驱鸟.为了解决上述问题,本文提出基于改进Q⁃learning算法的拟声驱鸟策略.首先,为了评价各音频的驱鸟效果,结合模糊理论,将鸟类听到音频后的动作行为量化为不同鸟类反应类型.然后,设计单一音频驱鸟实验,统计各音频驱鸟效果数据,得到各音频的初始权重值,为拟声驱鸟装置的音频选择提供实验依据.为了使计算所得的音频权重值更符合实际实验情况,对CRITIC(Criteria Impor⁃tance Though Intercrieria Correlation)方法的权重计算公式进行了优化.最后,使用实验所得音频权重值对Q⁃learning算法进行改进,并设计与其他拟声驱鸟策略的对比实验,实验数据显示改进Q⁃learning算法的拟声驱鸟策略驱鸟效果优于其他三种驱鸟策略,收敛速度快,驱鸟效果稳定,能够降低鸟类的适应性.展开更多
As weapon system effectiveness is affected by many factors,its evaluation is essentially a multi-criterion decision making problem for its complexity.The evaluation model of the effectiveness is established on the bas...As weapon system effectiveness is affected by many factors,its evaluation is essentially a multi-criterion decision making problem for its complexity.The evaluation model of the effectiveness is established on the basis of metrics architecture of the effectiveness.The Bayesian network,which is used to evaluate the effectiveness,is established based on the metrics architecture and the evaluation models.For getting the weights of the metrics by Bayesian network,subjective initial values of the weights are given,gradient ascent algorithm is adopted,and the reasonable values of the weights are achieved.And then the effectiveness of every weapon system project is gained.The weapon system,whose effectiveness is relative maximum,is the optimization system.The research result shows that this method can solve the problem of AHP method which evaluation results are not compatible to the practice results and overcome the shortcoming of neural network in multilayer and multi-criterion decision.The method offers a new approach for evaluating the effectiveness.展开更多
Brain functional networks model the brain's ability to exchange information across different regions,aiding in the understanding of the cognitive process of human visual attention during target searching,thereby c...Brain functional networks model the brain's ability to exchange information across different regions,aiding in the understanding of the cognitive process of human visual attention during target searching,thereby contributing to the advancement of camouflage evaluation.In this study,images with various camouflage effects were presented to observers to generate electroencephalography(EEG)signals,which were then used to construct a brain functional network.The topological parameters of the network were subsequently extracted and input into a machine learning model for training.The results indicate that most of the classifiers achieved accuracy rates exceeding 70%.Specifically,the Logistic algorithm achieved an accuracy of 81.67%.Therefore,it is possible to predict target camouflage effectiveness with high accuracy without the need to calculate discovery probability.The proposed method fully considers the aspects of human visual and cognitive processes,overcomes the subjectivity of human interpretation,and achieves stable and reliable accuracy.展开更多
For most open universities,in fact for most universities,there will arise occasions when some online activities will have to be completed by all students in a very short time interval.These occasions usually result in...For most open universities,in fact for most universities,there will arise occasions when some online activities will have to be completed by all students in a very short time interval.These occasions usually result in a great rush to get online as quickly as possible;the rush is often made worst by the limited good choices on offer such that usually only the earliest can get the best.Choosing and confirming elective subjects,when only a limited number for each are offered,on a first-come-first-served basis is an example. At the Open University Malaysia the rush occurs when confirming offered subjects and choosing face-to-face meeting timetable.Most students rushed to be earliest to ensure choice slots in the timetable are obtained. Available I.T.resources such as Internet bandwidth and servers cannot cope with these short-term sustained peak demands.Unless these peak demands are met,however,online services will slow down drastically resulting in long queues of users waiting to be served.Providing excess capacity,if at all possible,to ensure these short duration peak demands can be met would not be cost-effective since most of the time these resources are not utilised.This paper will look at some cost-effective approaches to meeting these short-term peak demands.展开更多
基金supported by National Natural Science Foundation of China(Grant No.12432018,12372346)the Innovative Research Groups of the National Natural Science Foundation of China(Grant No.12221002).
文摘A typical Whipple shield consists of double-layered plates with a certain gap.The space debris impacts the outer plate and is broken into a debris cloud(shattered,molten,vaporized)with dispersed energy and momentum,which reduces the risk of penetrating the bulkhead.In the realm of hypervelocity impact,strain rate(>10^(5)s^(-1))effects are negligible,and fluid dynamics is employed to describe the impact process.Efficient numerical tools for precisely predicting the damage degree can greatly accelerate the design and optimization of advanced protective structures.Current hypervelocity impact research primarily focuses on the interaction between projectile and front plate and the movement of debris cloud.However,the damage mechanism of debris cloud impacts on rear plates-the critical threat component-remains underexplored owing to complex multi-physics processes and prohibitive computational costs.Existing approaches,ranging from semi-empirical equations to a machine learningbased ballistic limit prediction method,are constrained to binary penetration classification.Alternatively,the uneven data from experiments and simulations caused these methods to be ineffective when the projectile has irregular shapes and complicate flight attitude.Therefore,it is urgent to develop a new damage prediction method for predicting the rear plate damage,which can help to gain a deeper understanding of the damage mechanism.In this study,a machine learning(ML)method is developed to predict the damage distribution in the rear plate.Based on the unit velocity space,the discretized information of debris cloud and rear plate damage from rare simulation cases is used as input data for training the ML models,while the generalization ability for damage distribution prediction is tested by other simulation cases with different attack angles.The results demonstrate that the training and prediction accuracies using the Random Forest(RF)algorithm significantly surpass those using Artificial Neural Networks(ANNs)and Support Vector Machine(SVM).The RF-based model effectively identifies damage features in sparsely distributed debris cloud and cumulative effect.This study establishes an expandable new dataset that accommodates additional parameters to improve the prediction accuracy.Results demonstrate the model's ability to overcome data imbalance limitations through debris cloud features,enabling rapid and accurate rear plate damage prediction across wider scenarios with minimal data requirements.
基金supported by the Anhui Postdoctoral Scientific Research Program Foundation(2022B579).
文摘Many digital platforms have employed free-content promotion strategies to deal with the high uncertainty levels regarding digital content products.However,the diversity of digital content products and user heterogeneity in content preference may blur the impact of platform promotions across users and products.Therefore,free-content promotion strategies should be adapted to allocate marketing resources optimally and increase revenue.This study develops personal-ized free-content promotion strategies based on individual-level heterogeneous treatment effects and explores the causes of their heterogeneity,focusing on the moderating effect of user engagement-related variables.To this end,we utilize ran-dom field experimental data provided by a top Chinese e-book platform.We employ a framework that combines machine learning with econometric causal inference methods to estimate individual treatment effects and analyze their potential mechanisms.The analysis shows that,on average,free-content promotions lead to a significant increase in consumer pay-ments.However,the higher the level of user engagement,the lower the payment lift caused by promotions,as more-engaged users are more strongly affected by the cannibalization effect of free-content promotion.This study introduces a novel causal research design to help platforms improve their marketing strategies.
基金supported by the National Natural Science Foundation of China(62171088,U19A2052,62020106011)the Medico-Engineering Cooperation Funds from University of Electronic Science and Technology of China(ZYGX2021YGLH215,ZYGX2022YGRH005)。
文摘Deep neural networks(DNNs)have achieved great success in many data processing applications.However,high computational complexity and storage cost make deep learning difficult to be used on resource-constrained devices,and it is not environmental-friendly with much power cost.In this paper,we focus on low-rank optimization for efficient deep learning techniques.In the space domain,DNNs are compressed by low rank approximation of the network parameters,which directly reduces the storage requirement with a smaller number of network parameters.In the time domain,the network parameters can be trained in a few subspaces,which enables efficient training for fast convergence.The model compression in the spatial domain is summarized into three categories as pre-train,pre-set,and compression-aware methods,respectively.With a series of integrable techniques discussed,such as sparse pruning,quantization,and entropy coding,we can ensemble them in an integration framework with lower computational complexity and storage.In addition to summary of recent technical advances,we have two findings for motivating future works.One is that the effective rank,derived from the Shannon entropy of the normalized singular values,outperforms other conventional sparse measures such as the?_1 norm for network compression.The other is a spatial and temporal balance for tensorized neural networks.For accelerating the training of tensorized neural networks,it is crucial to leverage redundancy for both model compression and subspace training.
基金Supported by the State Key Program of National Natural Science of China（70931001）, the Science Fund for Creative Research Group of National Natural Science Foundation of China （60821063）, National Science and Technology Support Plan of China （2006BAH02A09）, the Science Fund for Youth Scholars of Ministry of Education of China （200801451053）, and the Research Committee and the Department of Industrial and Systems Engineering of Hong Kong Polytechnic University Research Grants （G-U323）
基金the National Defense Science and Technology Key Laboratory Fund of China(XM2020XT1023).
文摘As one of the most important part of weapon system of systems(WSoS),quantitative evaluation of reconnaissance satellite system(RSS)is indispensable during its construction and application.Aiming at the problem of nonlinear effectiveness evaluation under small sample conditions,we propose an evaluation method based on support vector regression(SVR)to effectively address the defects of traditional methods.Considering the performance of SVR is influenced by the penalty factor,kernel type,and other parameters deeply,the improved grey wolf optimizer(IGWO)is employed for parameter optimization.In the proposed IGWO algorithm,the opposition-based learning strategy is adopted to increase the probability of avoiding the local optima,the mutation operator is used to escape from premature convergence and differential convergence factors are applied to increase the rate of convergence.Numerical experiments of 14 test functions validate the applicability of IGWO algorithm dealing with global optimization.The index system and evaluation method are constructed based on the characteristics of RSS.To validate the proposed IGWO-SVR evaluation method,eight benchmark data sets and combat simulation are employed to estimate the evaluation accuracy,convergence performance and computational complexity.According to the experimental results,the proposed method outperforms several prediction based evaluation methods,verifies the superiority and effectiveness in RSS operational effectiveness evaluation.
文摘日益频繁的鸟类活动给输电线路的安全运行带来了极大威胁,而现有拟声驱鸟装置由于缺乏智能性,无法长期有效驱鸟.为了解决上述问题,本文提出基于改进Q⁃learning算法的拟声驱鸟策略.首先,为了评价各音频的驱鸟效果,结合模糊理论,将鸟类听到音频后的动作行为量化为不同鸟类反应类型.然后,设计单一音频驱鸟实验,统计各音频驱鸟效果数据,得到各音频的初始权重值,为拟声驱鸟装置的音频选择提供实验依据.为了使计算所得的音频权重值更符合实际实验情况,对CRITIC(Criteria Impor⁃tance Though Intercrieria Correlation)方法的权重计算公式进行了优化.最后,使用实验所得音频权重值对Q⁃learning算法进行改进,并设计与其他拟声驱鸟策略的对比实验,实验数据显示改进Q⁃learning算法的拟声驱鸟策略驱鸟效果优于其他三种驱鸟策略,收敛速度快,驱鸟效果稳定,能够降低鸟类的适应性.
文摘As weapon system effectiveness is affected by many factors,its evaluation is essentially a multi-criterion decision making problem for its complexity.The evaluation model of the effectiveness is established on the basis of metrics architecture of the effectiveness.The Bayesian network,which is used to evaluate the effectiveness,is established based on the metrics architecture and the evaluation models.For getting the weights of the metrics by Bayesian network,subjective initial values of the weights are given,gradient ascent algorithm is adopted,and the reasonable values of the weights are achieved.And then the effectiveness of every weapon system project is gained.The weapon system,whose effectiveness is relative maximum,is the optimization system.The research result shows that this method can solve the problem of AHP method which evaluation results are not compatible to the practice results and overcome the shortcoming of neural network in multilayer and multi-criterion decision.The method offers a new approach for evaluating the effectiveness.
基金sponsored by the National Defense Science and Technology Key Laboratory Fund(Grant No.61422062205)the Equipment Pre-Research Fund(Grant No.JCKYS2022LD9)。
文摘Brain functional networks model the brain's ability to exchange information across different regions,aiding in the understanding of the cognitive process of human visual attention during target searching,thereby contributing to the advancement of camouflage evaluation.In this study,images with various camouflage effects were presented to observers to generate electroencephalography(EEG)signals,which were then used to construct a brain functional network.The topological parameters of the network were subsequently extracted and input into a machine learning model for training.The results indicate that most of the classifiers achieved accuracy rates exceeding 70%.Specifically,the Logistic algorithm achieved an accuracy of 81.67%.Therefore,it is possible to predict target camouflage effectiveness with high accuracy without the need to calculate discovery probability.The proposed method fully considers the aspects of human visual and cognitive processes,overcomes the subjectivity of human interpretation,and achieves stable and reliable accuracy.
文摘For most open universities,in fact for most universities,there will arise occasions when some online activities will have to be completed by all students in a very short time interval.These occasions usually result in a great rush to get online as quickly as possible;the rush is often made worst by the limited good choices on offer such that usually only the earliest can get the best.Choosing and confirming elective subjects,when only a limited number for each are offered,on a first-come-first-served basis is an example. At the Open University Malaysia the rush occurs when confirming offered subjects and choosing face-to-face meeting timetable.Most students rushed to be earliest to ensure choice slots in the timetable are obtained. Available I.T.resources such as Internet bandwidth and servers cannot cope with these short-term sustained peak demands.Unless these peak demands are met,however,online services will slow down drastically resulting in long queues of users waiting to be served.Providing excess capacity,if at all possible,to ensure these short duration peak demands can be met would not be cost-effective since most of the time these resources are not utilised.This paper will look at some cost-effective approaches to meeting these short-term peak demands.
