In differentiable search architecture search methods,a more efficient search space design can significantly improve the performance of the searched architecture,thus requiring people to carefully define the search spa...In differentiable search architecture search methods,a more efficient search space design can significantly improve the performance of the searched architecture,thus requiring people to carefully define the search space with different complexity according to various operations.Meanwhile rationalizing the search strategies to explore the well-defined search space will further improve the speed and efficiency of architecture search.With this in mind,we propose a faster and more efficient differentiable architecture search method,AllegroNAS.Firstly,we introduce a more efficient search space enriched by the introduction of two redefined convolution modules.Secondly,we utilize a more efficient architectural parameter regularization method,mitigating the overfitting problem during the search process and reducing the error brought about by gradient approximation.Meanwhile,we introduce a natural exponential cosine annealing method to make the learning rate of the neural network training process more suitable for the search procedure.Moreover,group convolution and data augmentation are employed to reduce the computational cost.Finally,through extensive experiments on several public datasets,we demonstrate that our method can more swiftly search for better-performing neural network architectures in a more efficient search space,thus validating the effectiveness of our approach.展开更多
In this paper, we use the global search characteristics of genetic algorithms to help search the weight space of the neurons in the cascade-correlation architecture. The cascade-correlation learning architecture is a ...In this paper, we use the global search characteristics of genetic algorithms to help search the weight space of the neurons in the cascade-correlation architecture. The cascade-correlation learning architecture is a technique of training and building neural networks that starts with a simple network of neurons and adds additional neurons as they are needed to suit a particular problem. In our approach, instead ofmodifying the genetic algorithm to account for convergence problems, we search the weight-space using the genetic algorithm and then apply the gradient technique of Quickprop to optimize the weights. This hybrid algorithm which is a combination of genetic algorithms and cascade-correlation is applied to the two spirals problem. We also use our algorithm in the prediction of the cyclic oxidation resistance of Ni- and Co-base superalloys.展开更多
An optimizing method of observation scheduling based on time-division multiplexing is proposed in this paper,and its efficiency is verified by outdoor experiments. The initial observation scheduling is first obtained ...An optimizing method of observation scheduling based on time-division multiplexing is proposed in this paper,and its efficiency is verified by outdoor experiments. The initial observation scheduling is first obtained by using a semi-random search algorithm,and secondly the connection time pair（ CTP） between adjacent objects is optimized by using a genetic algorithm. After obtaining these two parameters,the final observation scheduling can be obtained. According to pre-designed tracks between each adjacent objects in observation order,the seamless observation of neighboring targets is derived by automatically steering the antenna beam,so the observation efficiency is improved.展开更多
The optimization of process parameters in polyolefin production can bring significant economic benefits to the factory.However,due to small data sets,high costs associated with parameter verification cycles,and diffic...The optimization of process parameters in polyolefin production can bring significant economic benefits to the factory.However,due to small data sets,high costs associated with parameter verification cycles,and difficulty in establishing an optimization model,the optimization process is often restricted.To address this issue,we propose using a transfer learning Bayesian optimization strategy to improve the efficiency of parameter optimization while minimizing resource consumption.Specifically,we leverage Gaussian process(GP)regression models to establish an integrated model that incorporates both source and target grade production task data.We then measure the similarity weights of each model by comparing their predicted trends,and utilize these weights to accelerate the solution of optimal process parameters for producing target polyolefin grades.In order to enhance the accuracy of our approach,we acknowledge that measuring similarity in a global search space may not effectively capture local similarity characteristics.Therefore,we propose a novel method for transfer learning optimization that operates within a local space(LSTL-PBO).This method employs partial data acquired through random sampling from the target task data and utilizes Bayesian optimization techniques for model establishment.By focusing on a local search space,we aim to better discern and leverage the inherent similarities between source tasks and the target task.Additionally,we incorporate a parallel concept into our method to address multiple local search spaces simultaneously.By doing so,we can explore different regions of the parameter space in parallel,thereby increasing the chances of finding optimal process parameters.This localized approach allows us to improve the precision and effectiveness of our optimization process.The performance of our method is validated through experiments on benchmark problems,and we discuss the sensitivity of its hyperparameters.The results show that our proposed method can significantly improve the efficiency of process parameter optimization,reduce the dependence on source tasks,and enhance the method's robustness.This has great potential for optimizing processes in industrial environments.展开更多
Differential evolution （DE） is an evolutionary optimization method, which has been successfully used in many practical cases. However, DE involves large computation time, especially, when used to optimize the compur...Differential evolution （DE） is an evolutionary optimization method, which has been successfully used in many practical cases. However, DE involves large computation time, especially, when used to optimize the compurationally expensive objective function. To overcome this .difficulty, the concept of immunity based on vaccination is used to help proliferate excellent schemata and to restrain the degenerate phenomenon. To improve the effective- ness of vaccines, a new vaccine autonomous obtaining method, and a method of deciding the probability of vacci- nation are proposed. In addition, a method for modifying the search space dynamically is proposed to enhance the possibility of converging to the true global optimum. Experiments showed that the improved DE performs better than the classical DE significantly.展开更多
Ordering based search methods have advantages over graph based search methods for structure learning of Bayesian networks in terms on the efficiency. With the aim of further increasing the accuracy of ordering based s...Ordering based search methods have advantages over graph based search methods for structure learning of Bayesian networks in terms on the efficiency. With the aim of further increasing the accuracy of ordering based search methods, we first propose to increase the search space, which can facilitate escaping from the local optima. We present our search operators with majorizations, which are easy to implement. Experiments show that the proposed algorithm can obtain significantly more accurate results. With regard to the problem of the decrease on efficiency due to the increase of the search space, we then propose to add path priors as constraints into the swap process. We analyze the coefficient which may influence the performance of the proposed algorithm, the experiments show that the constraints can enhance the efficiency greatly, while has little effect on the accuracy. The final experiments show that, compared to other competitive methods, the proposed algorithm can find better solutions while holding high efficiency at the same time on both synthetic and real data sets.展开更多
A novel algorithm named randomized binary gravita- tional search （RBGS） algorithm is proposed for the set covering problem （SCP）. It differs from previous SCP approaches because it does not work directly on the SC...A novel algorithm named randomized binary gravita- tional search （RBGS） algorithm is proposed for the set covering problem （SCP）. It differs from previous SCP approaches because it does not work directly on the SCP matrix. In the proposed algo- rithm, the solution of SCP is viewed as multi-dimension position of objects in the binary search space. All objects in the space attract each other by the gravity force, and this force causes a global movement of all objects towards the objects with heavier masses which correspond to good solutions. Computation results show that the proposed algorithm is very competitive. In addition, the proposed aigodthm is extended for SCP to solve the fault diagno- sis problem in graph-based systems.展开更多
Minimizing network coding resources of multicast networks,such as the number of coding nodes or links,has been proved to be NP-hard,and taking propagation delay into account makes the problem more complicated. To reso...Minimizing network coding resources of multicast networks,such as the number of coding nodes or links,has been proved to be NP-hard,and taking propagation delay into account makes the problem more complicated. To resolve this optimal problem,an integer encoding routing-based genetic algorithm( REGA) is presented to map the optimization problem into a genetic algorithm( GA)framework. Moreover,to speed up the search process of the algorithm,an efficient local search procedure which can reduce the searching space size is designed for searching the feasible solution.Compared with the binary link state encoding representation genetic algorithm( BLSGA),the chromosome length of REGA is shorter and just depends on the number of sinks. Simulation results show the advantages of the algorithm in terms of getting the optimal solution and algorithmic convergence speed.展开更多
Image registration is the overlaying of two images of the same scene taken at different times or by different sensors. It is one of the essential steps in information processing in remote sensing. To attain a highly a...Image registration is the overlaying of two images of the same scene taken at different times or by different sensors. It is one of the essential steps in information processing in remote sensing. To attain a highly accurate, reliable and low computation cost in image registration a suitable and similarity metric and reduction in search data and search space is required. In this paper, the author shows that if the right bin size is chosen, mutual information can be more robust than correlation in the registration of multi-temporal images. The author also compares the sensitivity of mutual information and correlation to Gaussian and multiplicative speckle noise. The author investigates automatic subimage selection as a reduction in search data strategy. The author proposes a measure, called alienability, which shows the ability ofa subimage to provide reliable registration. Alternate subimage selection methods such as using gradient, entropy and variance are also investigated. The author furthermore looks into a search space strategy using a gradient approach to maximize mutual information and show our first results.展开更多
One of the important steps in mining event sequences is to find frequent episodes. Once the frequent episodes are discovered, rules about temporal relationships can he derived. In this paper, an cfficient algorithm fo...One of the important steps in mining event sequences is to find frequent episodes. Once the frequent episodes are discovered, rules about temporal relationships can he derived. In this paper, an cfficient algorithm for discovering frequent episodes is presented based on the level-wise search algorithm WINEPI. The proposed algorithm gains hetter candidate generation quality by introducing a new Lemma to help to target the combinations of episodes that are interesting in the next level and thins reduces the execution time. Experimental results on artificial and real data show the enhanced efficiency of the algorithm.展开更多
AutoML(Automated Machine Learning)is an emerging field that aims to automate the process of building machine learning models.AutoML emerged to increase productivity and efficiency by automating as much as possible the...AutoML(Automated Machine Learning)is an emerging field that aims to automate the process of building machine learning models.AutoML emerged to increase productivity and efficiency by automating as much as possible the inefficient work that occurs while repeating this process whenever machine learning is applied.In particular,research has been conducted for a long time on technologies that can effectively develop high-quality models by minimizing the intervention of model developers in the process from data preprocessing to algorithm selection and tuning.In this semantic review research,we summarize the data processing requirements for AutoML approaches and provide a detailed explanation.We place greater emphasis on neural architecture search(NAS)as it currently represents a highly popular sub-topic within the field of AutoML.NAS methods use machine learning algorithms to search through a large space of possible architectures and find the one that performs best on a given task.We provide a summary of the performance achieved by representative NAS algorithms on the CIFAR-10,CIFAR-100,ImageNet and wellknown benchmark datasets.Additionally,we delve into several noteworthy research directions in NAS methods including one/two-stage NAS,one-shot NAS and joint hyperparameter with architecture optimization.We discussed how the search space size and complexity in NAS can vary depending on the specific problem being addressed.To conclude,we examine several open problems(SOTA problems)within current AutoML methods that assure further investigation in future research.展开更多
The state space explosion problem is still the key obstacle for applying model checking to systems of industrial size. Abstraction-based methods have been particularly successful in this regard. This paper presents an...The state space explosion problem is still the key obstacle for applying model checking to systems of industrial size. Abstraction-based methods have been particularly successful in this regard. This paper presents an approach based on refinement of search space partition and abstraction which combines these two techniques for reducing the complexity of model checking. The refinement depends on the representation of each portion of search space. Especially, search space can be refined stepwise to get a better reduction. As reported in the case study, the integration of search space partition and abstraction improves the efficiency of verification with respect to the requirement of memory and obtains significant advantage over the use of each of them in isolation.展开更多
A multi-input multi-output(MIMO)detection scheme that requires considerable low complexity but still achieves the near optimal performance is proposed.The fundamental idea of the proposed MIMO detection scheme consist...A multi-input multi-output(MIMO)detection scheme that requires considerable low complexity but still achieves the near optimal performance is proposed.The fundamental idea of the proposed MIMO detection scheme consists of two points:1)the computational complexity is restrained by a complexity limit in low signal-to-noise ratio(SNR)region;2)while in high SNR region,the complexity is significantly reduced by the proposed search space method.Comparing with existing fixed-complexity techniques of MIMO detection(e.g.,K-best sphere detector and reduced-search maximum-likelihood(RS ML)detection),the significant benefit of proposed detection scheme is that less computational power will be spent for the given data rate,or the throughput of detector can be increased for high SNR cases.According to the simulation results,the near optimal performance can be obtained while the detection complexity is kept considerable small.展开更多
Large-scale multi-objective optimization problems(MOPs)that involve a large number of decision variables,have emerged from many real-world applications.While evolutionary algorithms(EAs)have been widely acknowledged a...Large-scale multi-objective optimization problems(MOPs)that involve a large number of decision variables,have emerged from many real-world applications.While evolutionary algorithms(EAs)have been widely acknowledged as a mainstream method for MOPs,most research progress and successful applications of EAs have been restricted to MOPs with small-scale decision variables.More recently,it has been reported that traditional multi-objective EAs(MOEAs)suffer severe deterioration with the increase of decision variables.As a result,and motivated by the emergence of real-world large-scale MOPs,investigation of MOEAs in this aspect has attracted much more attention in the past decade.This paper reviews the progress of evolutionary computation for large-scale multi-objective optimization from two angles.From the key difficulties of the large-scale MOPs,the scalability analysis is discussed by focusing on the performance of existing MOEAs and the challenges induced by the increase of the number of decision variables.From the perspective of methodology,the large-scale MOEAs are categorized into three classes and introduced respectively:divide and conquer based,dimensionality reduction based and enhanced search-based approaches.Several future research directions are also discussed.展开更多
We introduce a hybrid algorithm for the 01 multidimensional multi-objective knapsack problem. This algorithm, called GTS MOKP, combines a genetic procedure and a tabu search operator. The algorithm is evaluated on 9 ...We introduce a hybrid algorithm for the 01 multidimensional multi-objective knapsack problem. This algorithm, called GTS MOKP, combines a genetic procedure and a tabu search operator. The algorithm is evaluated on 9 well-known benchmark instances and shows highly competitive results compared with two state-of-the-art algorithms.展开更多
Partially observable Markov decision processes （POMDPs） provide a rich mathematical framework for planning tasks in partially observable stochastic environments. The notion of the covering number, a metric of captur...Partially observable Markov decision processes （POMDPs） provide a rich mathematical framework for planning tasks in partially observable stochastic environments. The notion of the covering number, a metric of capturing the search space size of a POMDP planning problem, has been proposed as a complexity measure of approximate POMDP planning. Existing theoretical results are based on POMDPs with finite and discrete state spaces and measured in the l1- metric space. When considering heuristics, they are assumed to be always admissible. This paper extends the theoretical results on the covering numbers of different search spaces, including the newly defined space reachable under inadmissible heuristics, to the ln-metric spaces. We provide a simple but scalable algorithm for estimating covering numbers. Experimentally, we provide estimated covering numbers of the search spaces reachable by following different policies on several benchmark problems, and analyze their abilities to predict the runtime of POMDP planning algorithms.展开更多
Purpose–The purpose of this paper is to provide an effective and simple technique to structural damage identification,particularly to identify a crack in a structure.Artificial neural networks approach is an alternat...Purpose–The purpose of this paper is to provide an effective and simple technique to structural damage identification,particularly to identify a crack in a structure.Artificial neural networks approach is an alternative to identify the extent and location of the damage over the classical methods.Radial basis function(RBF)networks are good at function mapping and generalization ability among the various neural network approaches.RBF neural networks are chosen for the present study of crack identification.Design/methodology/approach–Analyzing the vibration response of a structure is an effective way to monitor its health and even to detect the damage.A novel two-stage improved radial basis function(IRBF)neural network methodology with conventional RBF in the first stage and a reduced search space moving technique in the second stage is proposed to identify the crack in a cantilever beam structure in the frequency domain.Latin hypercube sampling(LHS)technique is used in both stages to sample the frequency modal patterns to train the proposed network.Study is also conducted with and without addition of 5%white noise to the input patterns to simulate the experimental errors.Findings–The results show a significant improvement in identifying the location and magnitude of a crack by the proposed IRBF method,in comparison with conventional RBF method and other classical methods.In case of crack location in a beam,the average identification error over 12 test cases was 0.69 per cent by IRBF network compared to 4.88 per cent by conventional RBF.Similar improvements are reported when compared to hybrid CPN BPN networks.It also requires much less computational effort as compared to other hybrid neural network approaches and classical methods.Originality/value–The proposed novel IRBF crack identification technique is unique in originality and not reported elsewhere.It can identify the crack location and crack depth with very good accuracy,less computational effort and ease of implementation.展开更多
Path-oriented test case generation is in essence a constraint satisfaction problem （CSP） solved by search strategies, among which backtracking algorithms are widely used. In this article, the backtracking algorithm ...Path-oriented test case generation is in essence a constraint satisfaction problem （CSP） solved by search strategies, among which backtracking algorithms are widely used. In this article, the backtracking algorithm branch and bound （BB） is introduced to generate path-oriented test cases automatically. A model based on state space search is proposed to construct the search tree dynamically. The BB is optimized from two perspectives. Variable permutation with a heuristic rule to break ties is adopted for the branching operation, and interval computation with analysis on the monotony of branching conditions is utilized for the bounding operation. Empirical experiments show that the proposed method performs well with linear complexity, and reaches 100% coverage on some benchmark programs with an advantage over some static and dynamic algorithms.展开更多
In this paper, we target a similarity search among data supply chains, which plays an essential role in optimizing the supply chain and extending its value. This problem is very challenging for application-oriented da...In this paper, we target a similarity search among data supply chains, which plays an essential role in optimizing the supply chain and extending its value. This problem is very challenging for application-oriented data supply chains because the high complexity of the data supply chain makes the computation of similarity extremely complex and inefficient. In this paper, we propose a feature space representation model based on key points,which can extract the key features from the subsequences of the original data supply chain and simplify it into a feature vector form. Then, we formulate the similarity computation of the subsequences based on the multiscale features. Further, we propose an improved hierarchical clustering algorithm for a similarity search over the data supply chains. The main idea is to separate the subsequences into disjoint groups such that each group meets one specific clustering criteria; thus, the cluster containing the query object is the similarity search result. The experimental results show that the proposed approach is both effective and efficient for data supply chain retrieval.展开更多
Although amazing progress has been made in ma- chine learning to achieve high generalization accuracy and ef- ficiency, there is still very limited work on deriving meaning- ful decision-making actions from the result...Although amazing progress has been made in ma- chine learning to achieve high generalization accuracy and ef- ficiency, there is still very limited work on deriving meaning- ful decision-making actions from the resulting models. How- ever, in many applications such as advertisement, recommen- dation systems, social networks, customer relationship man- agement, and clinical prediction, the users need not only ac- curate prediction, but also suggestions on actions to achieve a desirable goal （e.g., high ads hit rates） or avert an unde- sirable predicted result （e.g., clinical deterioration）. Existing works for extracting such actionability are few and limited to simple models such as a decision tree. The dilemma is that those models with high accuracy are often more complex and harder to extract actionability from. In this paper, we propose an effective method to extract ac- tionable knowledge from additive tree models （ATMs）, one of the most widely used and best off-the-shelf classifiers. We rigorously formulate the optimal actionable planning （OAP） problem for a given ATM, which is to extract an action- able plan for a given input so that it can achieve a desirable output while maximizing the net profit. Based on a state space graph formulation, we first propose an optimal heuris- tic search method which intends to find an optimal solution. Then, we also present a sub-optimal heuristic search with an admissible and consistent heuristic function which can re- markably improve the efficiency of the algorithm. Our exper- imental results demonstrate the effectiveness and efficiency of the proposed algorithms on several real datasets in the application domain of personal credit and banking.展开更多
基金This work was supported in part by the National Natural Science Foundation of China under Grant 61305001the Natural Science Foundation of Heilongjiang Province of China under Grant F201222.
文摘In differentiable search architecture search methods,a more efficient search space design can significantly improve the performance of the searched architecture,thus requiring people to carefully define the search space with different complexity according to various operations.Meanwhile rationalizing the search strategies to explore the well-defined search space will further improve the speed and efficiency of architecture search.With this in mind,we propose a faster and more efficient differentiable architecture search method,AllegroNAS.Firstly,we introduce a more efficient search space enriched by the introduction of two redefined convolution modules.Secondly,we utilize a more efficient architectural parameter regularization method,mitigating the overfitting problem during the search process and reducing the error brought about by gradient approximation.Meanwhile,we introduce a natural exponential cosine annealing method to make the learning rate of the neural network training process more suitable for the search procedure.Moreover,group convolution and data augmentation are employed to reduce the computational cost.Finally,through extensive experiments on several public datasets,we demonstrate that our method can more swiftly search for better-performing neural network architectures in a more efficient search space,thus validating the effectiveness of our approach.
文摘In this paper, we use the global search characteristics of genetic algorithms to help search the weight space of the neurons in the cascade-correlation architecture. The cascade-correlation learning architecture is a technique of training and building neural networks that starts with a simple network of neurons and adds additional neurons as they are needed to suit a particular problem. In our approach, instead ofmodifying the genetic algorithm to account for convergence problems, we search the weight-space using the genetic algorithm and then apply the gradient technique of Quickprop to optimize the weights. This hybrid algorithm which is a combination of genetic algorithms and cascade-correlation is applied to the two spirals problem. We also use our algorithm in the prediction of the cyclic oxidation resistance of Ni- and Co-base superalloys.
基金Supported by the National Natural Science Foundation of China(61271373,61571043)111 Project of China(B14010)
文摘An optimizing method of observation scheduling based on time-division multiplexing is proposed in this paper,and its efficiency is verified by outdoor experiments. The initial observation scheduling is first obtained by using a semi-random search algorithm,and secondly the connection time pair（ CTP） between adjacent objects is optimized by using a genetic algorithm. After obtaining these two parameters,the final observation scheduling can be obtained. According to pre-designed tracks between each adjacent objects in observation order,the seamless observation of neighboring targets is derived by automatically steering the antenna beam,so the observation efficiency is improved.
基金supported by National Natural Science Foundation of China(62394343)Major Program of Qingyuan Innovation Laboratory(00122002)+1 种基金Major Science and Technology Projects of Longmen Laboratory(231100220600)Shanghai Committee of Science and Technology(23ZR1416000)and Shanghai AI Lab.
文摘The optimization of process parameters in polyolefin production can bring significant economic benefits to the factory.However,due to small data sets,high costs associated with parameter verification cycles,and difficulty in establishing an optimization model,the optimization process is often restricted.To address this issue,we propose using a transfer learning Bayesian optimization strategy to improve the efficiency of parameter optimization while minimizing resource consumption.Specifically,we leverage Gaussian process(GP)regression models to establish an integrated model that incorporates both source and target grade production task data.We then measure the similarity weights of each model by comparing their predicted trends,and utilize these weights to accelerate the solution of optimal process parameters for producing target polyolefin grades.In order to enhance the accuracy of our approach,we acknowledge that measuring similarity in a global search space may not effectively capture local similarity characteristics.Therefore,we propose a novel method for transfer learning optimization that operates within a local space(LSTL-PBO).This method employs partial data acquired through random sampling from the target task data and utilizes Bayesian optimization techniques for model establishment.By focusing on a local search space,we aim to better discern and leverage the inherent similarities between source tasks and the target task.Additionally,we incorporate a parallel concept into our method to address multiple local search spaces simultaneously.By doing so,we can explore different regions of the parameter space in parallel,thereby increasing the chances of finding optimal process parameters.This localized approach allows us to improve the precision and effectiveness of our optimization process.The performance of our method is validated through experiments on benchmark problems,and we discuss the sensitivity of its hyperparameters.The results show that our proposed method can significantly improve the efficiency of process parameter optimization,reduce the dependence on source tasks,and enhance the method's robustness.This has great potential for optimizing processes in industrial environments.
基金Supported by the National Natural Science Foundation of China （60736021）, the National High Technology Research and Development Program of China （2006AA04Z184, 2007AA041406）, and the Key Technologies R＆D Program of Zhejiang Province （2006C 11066, 2006C31051）.
文摘Differential evolution （DE） is an evolutionary optimization method, which has been successfully used in many practical cases. However, DE involves large computation time, especially, when used to optimize the compurationally expensive objective function. To overcome this .difficulty, the concept of immunity based on vaccination is used to help proliferate excellent schemata and to restrain the degenerate phenomenon. To improve the effective- ness of vaccines, a new vaccine autonomous obtaining method, and a method of deciding the probability of vacci- nation are proposed. In addition, a method for modifying the search space dynamically is proposed to enhance the possibility of converging to the true global optimum. Experiments showed that the improved DE performs better than the classical DE significantly.
基金supported by the National Natural Science Fundation of China(61573285)the Doctoral Fundation of China(2013ZC53037)
文摘Ordering based search methods have advantages over graph based search methods for structure learning of Bayesian networks in terms on the efficiency. With the aim of further increasing the accuracy of ordering based search methods, we first propose to increase the search space, which can facilitate escaping from the local optima. We present our search operators with majorizations, which are easy to implement. Experiments show that the proposed algorithm can obtain significantly more accurate results. With regard to the problem of the decrease on efficiency due to the increase of the search space, we then propose to add path priors as constraints into the swap process. We analyze the coefficient which may influence the performance of the proposed algorithm, the experiments show that the constraints can enhance the efficiency greatly, while has little effect on the accuracy. The final experiments show that, compared to other competitive methods, the proposed algorithm can find better solutions while holding high efficiency at the same time on both synthetic and real data sets.
基金supported by the National Natural Science Foundation of China (4100605850909096)
文摘A novel algorithm named randomized binary gravita- tional search （RBGS） algorithm is proposed for the set covering problem （SCP）. It differs from previous SCP approaches because it does not work directly on the SCP matrix. In the proposed algo- rithm, the solution of SCP is viewed as multi-dimension position of objects in the binary search space. All objects in the space attract each other by the gravity force, and this force causes a global movement of all objects towards the objects with heavier masses which correspond to good solutions. Computation results show that the proposed algorithm is very competitive. In addition, the proposed aigodthm is extended for SCP to solve the fault diagno- sis problem in graph-based systems.
基金Supported by the National Natural Science Foundation of China(No.61473179)Shandong Province Higher Educational Science and Technology Program(No.J16LN20)+1 种基金Natural Science Foundation of Shandong Province(No.ZR2016FM18)the Youth Scholars Development Program of Shandong University of Technology
文摘Minimizing network coding resources of multicast networks,such as the number of coding nodes or links,has been proved to be NP-hard,and taking propagation delay into account makes the problem more complicated. To resolve this optimal problem,an integer encoding routing-based genetic algorithm( REGA) is presented to map the optimization problem into a genetic algorithm( GA)framework. Moreover,to speed up the search process of the algorithm,an efficient local search procedure which can reduce the searching space size is designed for searching the feasible solution.Compared with the binary link state encoding representation genetic algorithm( BLSGA),the chromosome length of REGA is shorter and just depends on the number of sinks. Simulation results show the advantages of the algorithm in terms of getting the optimal solution and algorithmic convergence speed.
文摘Image registration is the overlaying of two images of the same scene taken at different times or by different sensors. It is one of the essential steps in information processing in remote sensing. To attain a highly accurate, reliable and low computation cost in image registration a suitable and similarity metric and reduction in search data and search space is required. In this paper, the author shows that if the right bin size is chosen, mutual information can be more robust than correlation in the registration of multi-temporal images. The author also compares the sensitivity of mutual information and correlation to Gaussian and multiplicative speckle noise. The author investigates automatic subimage selection as a reduction in search data strategy. The author proposes a measure, called alienability, which shows the ability ofa subimage to provide reliable registration. Alternate subimage selection methods such as using gradient, entropy and variance are also investigated. The author furthermore looks into a search space strategy using a gradient approach to maximize mutual information and show our first results.
文摘One of the important steps in mining event sequences is to find frequent episodes. Once the frequent episodes are discovered, rules about temporal relationships can he derived. In this paper, an cfficient algorithm for discovering frequent episodes is presented based on the level-wise search algorithm WINEPI. The proposed algorithm gains hetter candidate generation quality by introducing a new Lemma to help to target the combinations of episodes that are interesting in the next level and thins reduces the execution time. Experimental results on artificial and real data show the enhanced efficiency of the algorithm.
文摘AutoML(Automated Machine Learning)is an emerging field that aims to automate the process of building machine learning models.AutoML emerged to increase productivity and efficiency by automating as much as possible the inefficient work that occurs while repeating this process whenever machine learning is applied.In particular,research has been conducted for a long time on technologies that can effectively develop high-quality models by minimizing the intervention of model developers in the process from data preprocessing to algorithm selection and tuning.In this semantic review research,we summarize the data processing requirements for AutoML approaches and provide a detailed explanation.We place greater emphasis on neural architecture search(NAS)as it currently represents a highly popular sub-topic within the field of AutoML.NAS methods use machine learning algorithms to search through a large space of possible architectures and find the one that performs best on a given task.We provide a summary of the performance achieved by representative NAS algorithms on the CIFAR-10,CIFAR-100,ImageNet and wellknown benchmark datasets.Additionally,we delve into several noteworthy research directions in NAS methods including one/two-stage NAS,one-shot NAS and joint hyperparameter with architecture optimization.We discussed how the search space size and complexity in NAS can vary depending on the specific problem being addressed.To conclude,we examine several open problems(SOTA problems)within current AutoML methods that assure further investigation in future research.
基金Supported by the National Natural Science Foundation of China (Grant Nos. 60573012 and 60421001) the National Grand FundamentalResearch 973 Program of China (Grant No. 2002cb312200)
文摘The state space explosion problem is still the key obstacle for applying model checking to systems of industrial size. Abstraction-based methods have been particularly successful in this regard. This paper presents an approach based on refinement of search space partition and abstraction which combines these two techniques for reducing the complexity of model checking. The refinement depends on the representation of each portion of search space. Especially, search space can be refined stepwise to get a better reduction. As reported in the case study, the integration of search space partition and abstraction improves the efficiency of verification with respect to the requirement of memory and obtains significant advantage over the use of each of them in isolation.
基金supported by the National Basic Research Program of China (No.2007CB310602).
文摘A multi-input multi-output(MIMO)detection scheme that requires considerable low complexity but still achieves the near optimal performance is proposed.The fundamental idea of the proposed MIMO detection scheme consists of two points:1)the computational complexity is restrained by a complexity limit in low signal-to-noise ratio(SNR)region;2)while in high SNR region,the complexity is significantly reduced by the proposed search space method.Comparing with existing fixed-complexity techniques of MIMO detection(e.g.,K-best sphere detector and reduced-search maximum-likelihood(RS ML)detection),the significant benefit of proposed detection scheme is that less computational power will be spent for the given data rate,or the throughput of detector can be increased for high SNR cases.According to the simulation results,the near optimal performance can be obtained while the detection complexity is kept considerable small.
基金This work was supported by the Natural Science Foundation of China(Nos.61672478 and 61806090)the National Key Research and Development Program of China(No.2017YFB1003102)+4 种基金the Guangdong Provincial Key Laboratory(No.2020B121201001)the Shenzhen Peacock Plan(No.KQTD2016112514355531)the Guangdong-Hong Kong-Macao Greater Bay Area Center for Brain Science and Brain-inspired Intelligence Fund(No.2019028)the Fellowship of China Postdoctoral Science Foundation(No.2020M671900)the National Leading Youth Talent Support Program of China.
文摘Large-scale multi-objective optimization problems(MOPs)that involve a large number of decision variables,have emerged from many real-world applications.While evolutionary algorithms(EAs)have been widely acknowledged as a mainstream method for MOPs,most research progress and successful applications of EAs have been restricted to MOPs with small-scale decision variables.More recently,it has been reported that traditional multi-objective EAs(MOEAs)suffer severe deterioration with the increase of decision variables.As a result,and motivated by the emergence of real-world large-scale MOPs,investigation of MOEAs in this aspect has attracted much more attention in the past decade.This paper reviews the progress of evolutionary computation for large-scale multi-objective optimization from two angles.From the key difficulties of the large-scale MOPs,the scalability analysis is discussed by focusing on the performance of existing MOEAs and the challenges induced by the increase of the number of decision variables.From the perspective of methodology,the large-scale MOEAs are categorized into three classes and introduced respectively:divide and conquer based,dimensionality reduction based and enhanced search-based approaches.Several future research directions are also discussed.
文摘We introduce a hybrid algorithm for the 01 multidimensional multi-objective knapsack problem. This algorithm, called GTS MOKP, combines a genetic procedure and a tabu search operator. The algorithm is evaluated on 9 well-known benchmark instances and shows highly competitive results compared with two state-of-the-art algorithms.
文摘Partially observable Markov decision processes （POMDPs） provide a rich mathematical framework for planning tasks in partially observable stochastic environments. The notion of the covering number, a metric of capturing the search space size of a POMDP planning problem, has been proposed as a complexity measure of approximate POMDP planning. Existing theoretical results are based on POMDPs with finite and discrete state spaces and measured in the l1- metric space. When considering heuristics, they are assumed to be always admissible. This paper extends the theoretical results on the covering numbers of different search spaces, including the newly defined space reachable under inadmissible heuristics, to the ln-metric spaces. We provide a simple but scalable algorithm for estimating covering numbers. Experimentally, we provide estimated covering numbers of the search spaces reachable by following different policies on several benchmark problems, and analyze their abilities to predict the runtime of POMDP planning algorithms.
文摘Purpose–The purpose of this paper is to provide an effective and simple technique to structural damage identification,particularly to identify a crack in a structure.Artificial neural networks approach is an alternative to identify the extent and location of the damage over the classical methods.Radial basis function(RBF)networks are good at function mapping and generalization ability among the various neural network approaches.RBF neural networks are chosen for the present study of crack identification.Design/methodology/approach–Analyzing the vibration response of a structure is an effective way to monitor its health and even to detect the damage.A novel two-stage improved radial basis function(IRBF)neural network methodology with conventional RBF in the first stage and a reduced search space moving technique in the second stage is proposed to identify the crack in a cantilever beam structure in the frequency domain.Latin hypercube sampling(LHS)technique is used in both stages to sample the frequency modal patterns to train the proposed network.Study is also conducted with and without addition of 5%white noise to the input patterns to simulate the experimental errors.Findings–The results show a significant improvement in identifying the location and magnitude of a crack by the proposed IRBF method,in comparison with conventional RBF method and other classical methods.In case of crack location in a beam,the average identification error over 12 test cases was 0.69 per cent by IRBF network compared to 4.88 per cent by conventional RBF.Similar improvements are reported when compared to hybrid CPN BPN networks.It also requires much less computational effort as compared to other hybrid neural network approaches and classical methods.Originality/value–The proposed novel IRBF crack identification technique is unique in originality and not reported elsewhere.It can identify the crack location and crack depth with very good accuracy,less computational effort and ease of implementation.
基金supported by the Hi-Tech Research and Development Program of China(2012AA011201)the National Natural Science Foundation of China(61202080)+1 种基金the Major Program of the National Natural Science Foundation of China(91318301)the Open Funding of State Key Laboratory of Computer Architecture(CARCH201201)
文摘Path-oriented test case generation is in essence a constraint satisfaction problem （CSP） solved by search strategies, among which backtracking algorithms are widely used. In this article, the backtracking algorithm branch and bound （BB） is introduced to generate path-oriented test cases automatically. A model based on state space search is proposed to construct the search tree dynamically. The BB is optimized from two perspectives. Variable permutation with a heuristic rule to break ties is adopted for the branching operation, and interval computation with analysis on the monotony of branching conditions is utilized for the bounding operation. Empirical experiments show that the proposed method performs well with linear complexity, and reaches 100% coverage on some benchmark programs with an advantage over some static and dynamic algorithms.
基金partly supported by the National Natural Science Foundation of China(Nos.61532012,61370196,and 61672109)
文摘In this paper, we target a similarity search among data supply chains, which plays an essential role in optimizing the supply chain and extending its value. This problem is very challenging for application-oriented data supply chains because the high complexity of the data supply chain makes the computation of similarity extremely complex and inefficient. In this paper, we propose a feature space representation model based on key points,which can extract the key features from the subsequences of the original data supply chain and simplify it into a feature vector form. Then, we formulate the similarity computation of the subsequences based on the multiscale features. Further, we propose an improved hierarchical clustering algorithm for a similarity search over the data supply chains. The main idea is to separate the subsequences into disjoint groups such that each group meets one specific clustering criteria; thus, the cluster containing the query object is the similarity search result. The experimental results show that the proposed approach is both effective and efficient for data supply chain retrieval.
基金This work was supported in part by China Postdoctoral Science Foundation （2013M531527）, the Fundamental Research Funds for the Central Universities （0110000037）, the National Natural Science Foun- dation of China （Grant Nos. 61502412, 61033009, and 61175057）, Natural Science Foundation of the Jiangsu Province （BK20150459）, Natural Science Foundation of the Jiangsu Higher Education Institutions （15KJB520036）, National Science Foundation, United States （IIS-0534699, IIS-0713109, CNS-1017701）, and a Microsoft Research New Faculty Fellowship.
文摘Although amazing progress has been made in ma- chine learning to achieve high generalization accuracy and ef- ficiency, there is still very limited work on deriving meaning- ful decision-making actions from the resulting models. How- ever, in many applications such as advertisement, recommen- dation systems, social networks, customer relationship man- agement, and clinical prediction, the users need not only ac- curate prediction, but also suggestions on actions to achieve a desirable goal （e.g., high ads hit rates） or avert an unde- sirable predicted result （e.g., clinical deterioration）. Existing works for extracting such actionability are few and limited to simple models such as a decision tree. The dilemma is that those models with high accuracy are often more complex and harder to extract actionability from. In this paper, we propose an effective method to extract ac- tionable knowledge from additive tree models （ATMs）, one of the most widely used and best off-the-shelf classifiers. We rigorously formulate the optimal actionable planning （OAP） problem for a given ATM, which is to extract an action- able plan for a given input so that it can achieve a desirable output while maximizing the net profit. Based on a state space graph formulation, we first propose an optimal heuris- tic search method which intends to find an optimal solution. Then, we also present a sub-optimal heuristic search with an admissible and consistent heuristic function which can re- markably improve the efficiency of the algorithm. Our exper- imental results demonstrate the effectiveness and efficiency of the proposed algorithms on several real datasets in the application domain of personal credit and banking.
