In order to meet the demand of online optimal running, a novel soft sensor modeling approach based on Gaussian processes was proposed. The approach is moderately simple to implement and use without loss of performance...In order to meet the demand of online optimal running, a novel soft sensor modeling approach based on Gaussian processes was proposed. The approach is moderately simple to implement and use without loss of performance. It is trained by optimizing the hyperparameters using the scaled conjugate gradient algorithm with the squared exponential covariance function employed. Experimental simulations show that the soft sensor modeling approach has the advantage via a real-world example in a refinery. Meanwhile, the method opens new possibilities for application of kernel methods to potential fields.展开更多
An improved approach for J-value segmentation （JSEG） is presented for unsupervised color image segmentation. Instead of color quantization algorithm, an automatic classification method based on adaptive mean shift ...An improved approach for J-value segmentation （JSEG） is presented for unsupervised color image segmentation. Instead of color quantization algorithm, an automatic classification method based on adaptive mean shift （AMS） based clustering is used for nonparametric clustering of image data set. The clustering results are used to construct Gaussian mixture modelling （GMM） of image data for the calculation of soft J value. The region growing algorithm used in JSEG is then applied in segmenting the image based on the multiscale soft J-images. Experiments show that the synergism of JSEG and the soft classification based on AMS based clustering and GMM overcomes the limitations of JSEG successfully and is more robust.展开更多
Since the joint probabilistic data association(JPDA)algorithm results in calculation explosion with the increasing number of targets,a multi-target tracking algorithm based on Gaussian mixture model(GMM)clustering is ...Since the joint probabilistic data association(JPDA)algorithm results in calculation explosion with the increasing number of targets,a multi-target tracking algorithm based on Gaussian mixture model(GMM)clustering is proposed.The algorithm is used to cluster the measurements,and the association matrix between measurements and tracks is constructed by the posterior probability.Compared with the traditional data association algorithm,this algorithm has better tracking performance and less computational complexity.Simulation results demonstrate the effectiveness of the proposed algorithm.展开更多
An efficient approach was proposed for discriminating shadows from moving objects. In the background subtraction stage, moving objects were extracted. Then, the initial classification for moving shadow pixels and fore...An efficient approach was proposed for discriminating shadows from moving objects. In the background subtraction stage, moving objects were extracted. Then, the initial classification for moving shadow pixels and foreground object pixels was performed by using color invariant features. In the shadow model learning stage, instead of a single Gaussian distribution, it was assumed that the density function computed on the values of chromaticity difference or bright difference, can be modeled as a mixture of Gaussian consisting of two density functions. Meanwhile, the Gaussian parameter estimation was performed by using EM algorithm. The estimates were used to obtain shadow mask according to two constraints. Finally, experiments were carried out. The visual experiment results confirm the effectiveness of proposed method. Quantitative results in terms of the shadow detection rate and the shadow discrimination rate(the maximum values are 85.79% and 97.56%, respectively) show that the proposed approach achieves a satisfying result with post-processing step.展开更多
The key problem of the adaptive mixture background model is that the parameters can adaptively change according to the input data. To address the problem, a new method is proposed. Firstly, the recursive equations are...The key problem of the adaptive mixture background model is that the parameters can adaptively change according to the input data. To address the problem, a new method is proposed. Firstly, the recursive equations are inferred based on the maximum likelihood rule. Secondly, the forgetting factor and learning rate factor are redefined, and their still more general formulations are obtained by analyzing their practical functions. Lastly, the convergence of the proposed algorithm is proved to enable the estimation converge to a local maximum of the data likelihood function according to the stochastic approximation theory. The experiments show that the proposed learning algorithm excels the formers both in converging rate and accuracy.展开更多
Gaussian process(GP)has fewer parameters,simple model and output of probabilistic sense,when compared with the methods such as support vector machines.Selection of the hyper-parameters is critical to the performance o...Gaussian process(GP)has fewer parameters,simple model and output of probabilistic sense,when compared with the methods such as support vector machines.Selection of the hyper-parameters is critical to the performance of Gaussian process model.However,the common-used algorithm has the disadvantages of difficult determination of iteration steps,over-dependence of optimization effect on initial values,and easily falling into local optimum.To solve this problem,a method combining the Gaussian process with memetic algorithm was proposed.Based on this method,memetic algorithm was used to search the optimal hyper parameters of Gaussian process regression(GPR)model in the training process and form MA-GPR algorithms,and then the model was used to predict and test the results.When used in the marine long-range precision strike system(LPSS)battle effectiveness evaluation,the proposed MA-GPR model significantly improved the prediction accuracy,compared with the conjugate gradient method and the genetic algorithm optimization process.展开更多
In order to reduce the computation of complex problems, a new surrogate-assisted estimation of distribution algorithm with Gaussian process was proposed. Coevolution was used in dual populations which evolved in paral...In order to reduce the computation of complex problems, a new surrogate-assisted estimation of distribution algorithm with Gaussian process was proposed. Coevolution was used in dual populations which evolved in parallel. The search space was projected into multiple subspaces and searched by sub-populations. Also, the whole space was exploited by the other population which exchanges information with the sub-populations. In order to make the evolutionary course efficient, multivariate Gaussian model and Gaussian mixture model were used in both populations separately to estimate the distribution of individuals and reproduce new generations. For the surrogate model, Gaussian process was combined with the algorithm which predicted variance of the predictions. The results on six benchmark functions show that the new algorithm performs better than other surrogate-model based algorithms and the computation complexity is only 10% of the original estimation of distribution algorithm.展开更多
为了解决联邦学习在车联网中终端设备数据的异质性导致模型训练准确率不稳定和性能下降,以及车辆分布广泛,通信和计算资源有限的问题,提出一种数据类型和数据规模并行优化的群联邦迁移学习数据共享方法(swarm federated transfer learni...为了解决联邦学习在车联网中终端设备数据的异质性导致模型训练准确率不稳定和性能下降,以及车辆分布广泛,通信和计算资源有限的问题,提出一种数据类型和数据规模并行优化的群联邦迁移学习数据共享方法(swarm federated transfer learning,SFTL)。提出基于高斯混合模型的共识设备组划分机制,通过对数据分布建模构建共识设备组,实现对异质性数据的有效管理和分析;面向划分的共识设备组,设计蜂群学习训练机制,加强相似设备组之间的协同学习过程;提出组间迁移学习机制,通过模型预训练法增量迁移不同共识设备组信息最小化模型差异,提高联邦模型聚合准确率。在公共数据集上的实验表明:与基线方法相比,SFTL模型训练准确率平均提高7%,通信时间平均降低10%。展开更多
文摘In order to meet the demand of online optimal running, a novel soft sensor modeling approach based on Gaussian processes was proposed. The approach is moderately simple to implement and use without loss of performance. It is trained by optimizing the hyperparameters using the scaled conjugate gradient algorithm with the squared exponential covariance function employed. Experimental simulations show that the soft sensor modeling approach has the advantage via a real-world example in a refinery. Meanwhile, the method opens new possibilities for application of kernel methods to potential fields.
文摘An improved approach for J-value segmentation （JSEG） is presented for unsupervised color image segmentation. Instead of color quantization algorithm, an automatic classification method based on adaptive mean shift （AMS） based clustering is used for nonparametric clustering of image data set. The clustering results are used to construct Gaussian mixture modelling （GMM） of image data for the calculation of soft J value. The region growing algorithm used in JSEG is then applied in segmenting the image based on the multiscale soft J-images. Experiments show that the synergism of JSEG and the soft classification based on AMS based clustering and GMM overcomes the limitations of JSEG successfully and is more robust.
基金the National Natural Science Foundation of China(61771367)the Science and Technology on Communication Networks Laboratory(HHS19641X003).
文摘Since the joint probabilistic data association(JPDA)algorithm results in calculation explosion with the increasing number of targets,a multi-target tracking algorithm based on Gaussian mixture model(GMM)clustering is proposed.The algorithm is used to cluster the measurements,and the association matrix between measurements and tracks is constructed by the posterior probability.Compared with the traditional data association algorithm,this algorithm has better tracking performance and less computational complexity.Simulation results demonstrate the effectiveness of the proposed algorithm.
基金Project(50805023)supported by the National Natural Science Foundation of ChinaProject(BA2010093)supported by the Special Fund of Jiangsu Province for the Transformation of Scientific and Technological Achievements,ChinaProject(2008144)supported by the Hexa-type Elites Peak Program of Jiangsu Province,China
文摘An efficient approach was proposed for discriminating shadows from moving objects. In the background subtraction stage, moving objects were extracted. Then, the initial classification for moving shadow pixels and foreground object pixels was performed by using color invariant features. In the shadow model learning stage, instead of a single Gaussian distribution, it was assumed that the density function computed on the values of chromaticity difference or bright difference, can be modeled as a mixture of Gaussian consisting of two density functions. Meanwhile, the Gaussian parameter estimation was performed by using EM algorithm. The estimates were used to obtain shadow mask according to two constraints. Finally, experiments were carried out. The visual experiment results confirm the effectiveness of proposed method. Quantitative results in terms of the shadow detection rate and the shadow discrimination rate(the maximum values are 85.79% and 97.56%, respectively) show that the proposed approach achieves a satisfying result with post-processing step.
基金the Doctorate Foundation of the Engineering College, Air Force Engineering University.
文摘The key problem of the adaptive mixture background model is that the parameters can adaptively change according to the input data. To address the problem, a new method is proposed. Firstly, the recursive equations are inferred based on the maximum likelihood rule. Secondly, the forgetting factor and learning rate factor are redefined, and their still more general formulations are obtained by analyzing their practical functions. Lastly, the convergence of the proposed algorithm is proved to enable the estimation converge to a local maximum of the data likelihood function according to the stochastic approximation theory. The experiments show that the proposed learning algorithm excels the formers both in converging rate and accuracy.
基金Project(513300303)supported by the General Armament Department,China
文摘Gaussian process(GP)has fewer parameters,simple model and output of probabilistic sense,when compared with the methods such as support vector machines.Selection of the hyper-parameters is critical to the performance of Gaussian process model.However,the common-used algorithm has the disadvantages of difficult determination of iteration steps,over-dependence of optimization effect on initial values,and easily falling into local optimum.To solve this problem,a method combining the Gaussian process with memetic algorithm was proposed.Based on this method,memetic algorithm was used to search the optimal hyper parameters of Gaussian process regression(GPR)model in the training process and form MA-GPR algorithms,and then the model was used to predict and test the results.When used in the marine long-range precision strike system(LPSS)battle effectiveness evaluation,the proposed MA-GPR model significantly improved the prediction accuracy,compared with the conjugate gradient method and the genetic algorithm optimization process.
基金Project(2009CB320603)supported by the National Basic Research Program of ChinaProject(IRT0712)supported by Program for Changjiang Scholars and Innovative Research Team in University+1 种基金Project(B504)supported by the Shanghai Leading Academic Discipline ProgramProject(61174118)supported by the National Natural Science Foundation of China
文摘In order to reduce the computation of complex problems, a new surrogate-assisted estimation of distribution algorithm with Gaussian process was proposed. Coevolution was used in dual populations which evolved in parallel. The search space was projected into multiple subspaces and searched by sub-populations. Also, the whole space was exploited by the other population which exchanges information with the sub-populations. In order to make the evolutionary course efficient, multivariate Gaussian model and Gaussian mixture model were used in both populations separately to estimate the distribution of individuals and reproduce new generations. For the surrogate model, Gaussian process was combined with the algorithm which predicted variance of the predictions. The results on six benchmark functions show that the new algorithm performs better than other surrogate-model based algorithms and the computation complexity is only 10% of the original estimation of distribution algorithm.
文摘为了解决联邦学习在车联网中终端设备数据的异质性导致模型训练准确率不稳定和性能下降,以及车辆分布广泛,通信和计算资源有限的问题,提出一种数据类型和数据规模并行优化的群联邦迁移学习数据共享方法(swarm federated transfer learning,SFTL)。提出基于高斯混合模型的共识设备组划分机制,通过对数据分布建模构建共识设备组,实现对异质性数据的有效管理和分析;面向划分的共识设备组,设计蜂群学习训练机制,加强相似设备组之间的协同学习过程;提出组间迁移学习机制,通过模型预训练法增量迁移不同共识设备组信息最小化模型差异,提高联邦模型聚合准确率。在公共数据集上的实验表明:与基线方法相比,SFTL模型训练准确率平均提高7%,通信时间平均降低10%。
