There have been vast amount of studies on background modeling to detect moving objects. Two recent reviews[1,2] showed that kernel density estimation(KDE) method and Gaussian mixture model(GMM) perform about equally b...There have been vast amount of studies on background modeling to detect moving objects. Two recent reviews[1,2] showed that kernel density estimation(KDE) method and Gaussian mixture model(GMM) perform about equally best among possible background models. For KDE, the selection of kernel functions and their bandwidths greatly influence the performance. There were few attempts to compare the adequacy of functions for KDE. In this paper, we evaluate the performance of various functions for KDE. Functions tested include almost everyone cited in the literature and a new function, Laplacian of Gaussian(LoG) is also introduced for comparison. All tests were done on real videos with vary-ing background dynamics and results were analyzed both qualitatively and quantitatively. Effect of different bandwidths was also investigated.展开更多
Upscaling of primary geological models with huge cells, especially in porous media, is the first step in fluid flow simulation. Numerical methods are often used to solve the models. The upscaling method must preserve ...Upscaling of primary geological models with huge cells, especially in porous media, is the first step in fluid flow simulation. Numerical methods are often used to solve the models. The upscaling method must preserve the important properties of the spatial distribution of the reservoir properties. An grid upscaling method based on adaptive bandwidth in kernel function is proposed according to the spatial distribution of property. This type of upscaling reduces the number of cells, while preserves the main heterogeneity features of the original fine model. The key point of the paper is upscaling two reservoir properties simultaneously. For each reservoir feature, the amount of bandwidth or optimal threshold is calculated and the results of the upscaling are obtained. Then two approaches are used to upscaling two properties simultaneously based on maximum bandwidth and minimum bandwidth. In fact, we now have a finalized upscaled model for both reservoir properties for each approach in which not only the number of their cells, but also the locations of the cells are equal. The upscaling error of the minimum bandwidth approach is less than that of the maximum bandwidth approach.展开更多
The uniform mathematical model of distortion signals in power grid has been setup with the theory of Wiener-G Functional. Firstly,the Matlab simulation models were established. Secondly,the Wiener kernel of power load...The uniform mathematical model of distortion signals in power grid has been setup with the theory of Wiener-G Functional. Firstly,the Matlab simulation models were established. Secondly,the Wiener kernel of power load was found based on the Gaussian white noise as input. And then the uniform mathematical model of the power grid signal was established according to the homogeneous of the same order of Wiener functional series. Finally,taking three typical distortion sources which are semiconductor rectifier,electric locomotive and electric arc furnace in power grid as examples,we have validated the model through the Matlab simulation and analyzed the simulation errors. The results show that the uniform mathematical model of distortion signals in power grid can approximation the actual model by growing the items of the series under the condition of the enough storage space and computing speed.展开更多
The development of forecasting models for pollution particles shows a nonlinear dynamic behavior;hence, implementation is a non-trivial process. In the literature, there have been multiple models of particulate pollut...The development of forecasting models for pollution particles shows a nonlinear dynamic behavior;hence, implementation is a non-trivial process. In the literature, there have been multiple models of particulate pollutants, which use softcomputing techniques and machine learning such as: multilayer perceptrons, neural networks, support vector machines, kernel algorithms, and so on. This paper presents a prediction pollution model using support vector machines and kernel functions, which are: Gaussian, Polynomial and Spline. Finally, the prediction results of ozone (O3), particulate matter (PM10) and nitrogen dioxide (NO2) at Mexico City are presented as a case study using these techniques.展开更多
通过融合多个核函数,提出一种多核主成分分析(multi-kernel principal component analysis,MKPCA)和二元Logistic回归耦合的诊断方法(MKPCA-Logistic回归模型)诊断冠心病,较好的解决了单一核函数适应性问题。选取第一舒张波高度U_(1)、...通过融合多个核函数,提出一种多核主成分分析(multi-kernel principal component analysis,MKPCA)和二元Logistic回归耦合的诊断方法(MKPCA-Logistic回归模型)诊断冠心病,较好的解决了单一核函数适应性问题。选取第一舒张波高度U_(1)、第三舒张波高度U_(3)、第一收缩波高度D_(1)、第二收缩波高度D_(2)、第三收缩波高度D_(3)、收缩波的波动值±U_(1)等6个影响因子,建立Logistic回归模型以及MKPCA-Logistic回归模型对冠心病进行诊断。利用预测准确率、误判率和成功率曲线(receiver operating characteristic,ROC)对两种模型的预测精度进行检验。结果表明:MKPCA-Logistic回归模型预测患冠心病的正确率为97%,明显高于Logistic回归模型的正确率92.5%。从ROC曲线分析来看,Logistic回归模型的ROC曲线的曲线下面积(AUC)为0.783,MKPCA-Logistic回归模型的AUC为0.874,耦合模型的分类精度更高。展开更多
文摘There have been vast amount of studies on background modeling to detect moving objects. Two recent reviews[1,2] showed that kernel density estimation(KDE) method and Gaussian mixture model(GMM) perform about equally best among possible background models. For KDE, the selection of kernel functions and their bandwidths greatly influence the performance. There were few attempts to compare the adequacy of functions for KDE. In this paper, we evaluate the performance of various functions for KDE. Functions tested include almost everyone cited in the literature and a new function, Laplacian of Gaussian(LoG) is also introduced for comparison. All tests were done on real videos with vary-ing background dynamics and results were analyzed both qualitatively and quantitatively. Effect of different bandwidths was also investigated.
文摘Upscaling of primary geological models with huge cells, especially in porous media, is the first step in fluid flow simulation. Numerical methods are often used to solve the models. The upscaling method must preserve the important properties of the spatial distribution of the reservoir properties. An grid upscaling method based on adaptive bandwidth in kernel function is proposed according to the spatial distribution of property. This type of upscaling reduces the number of cells, while preserves the main heterogeneity features of the original fine model. The key point of the paper is upscaling two reservoir properties simultaneously. For each reservoir feature, the amount of bandwidth or optimal threshold is calculated and the results of the upscaling are obtained. Then two approaches are used to upscaling two properties simultaneously based on maximum bandwidth and minimum bandwidth. In fact, we now have a finalized upscaled model for both reservoir properties for each approach in which not only the number of their cells, but also the locations of the cells are equal. The upscaling error of the minimum bandwidth approach is less than that of the maximum bandwidth approach.
基金Sponsored by the National Natural Science Foundation of China(Grant No.51277043)
文摘The uniform mathematical model of distortion signals in power grid has been setup with the theory of Wiener-G Functional. Firstly,the Matlab simulation models were established. Secondly,the Wiener kernel of power load was found based on the Gaussian white noise as input. And then the uniform mathematical model of the power grid signal was established according to the homogeneous of the same order of Wiener functional series. Finally,taking three typical distortion sources which are semiconductor rectifier,electric locomotive and electric arc furnace in power grid as examples,we have validated the model through the Matlab simulation and analyzed the simulation errors. The results show that the uniform mathematical model of distortion signals in power grid can approximation the actual model by growing the items of the series under the condition of the enough storage space and computing speed.
文摘The development of forecasting models for pollution particles shows a nonlinear dynamic behavior;hence, implementation is a non-trivial process. In the literature, there have been multiple models of particulate pollutants, which use softcomputing techniques and machine learning such as: multilayer perceptrons, neural networks, support vector machines, kernel algorithms, and so on. This paper presents a prediction pollution model using support vector machines and kernel functions, which are: Gaussian, Polynomial and Spline. Finally, the prediction results of ozone (O3), particulate matter (PM10) and nitrogen dioxide (NO2) at Mexico City are presented as a case study using these techniques.
