Enhanced kernel minimum squared error algorithm and its application in face recognition

To improve the classification performance of the kernel minimum squared error（ KMSE）, an enhanced KMSE algorithm（ EKMSE） is proposed. It redefines the regular objective function by introducing a novel class label definition, and the relative class label matrix can be adaptively adjusted to the kernel matrix.Compared with the common methods, the newobjective function can enlarge the distance between different classes, which therefore yields better recognition rates. In addition, an iteration parameter searching technique is adopted to improve the computational efficiency. The extensive experiments on FERET and GT face databases illustrate the feasibility and efficiency of the proposed EKMSE. It outperforms the original MSE, KMSE,some KMSE improvement methods, and even the sparse representation-based techniques in face recognition, such as collaborate representation classification（ CRC）.

基于改进K-means的局部离群点检测方法

离群点检测任务是指检测与正常数据在特征属性上存在显著差异的异常数据。大多数基于聚类的离群点检测方法主要从全局角度对数据集中的离群点进行检测,而对局部离群点的检测性能较弱。基于此,本文通过引入快速搜索和发现密度峰值方法改进K-means聚类算法,提出了一种名为KLOD(local outlier detection based on improved K-means and least-squares methods)的局部离群点检测方法,以实现对局部离群点的精确检测。首先,利用快速搜索和发现密度峰值方法计算数据点的局部密度和相对距离,并将二者相乘得到γ值。其次,将γ值降序排序,利用肘部法则选择γ值最大的k个数据点作为K-means聚类算法的初始聚类中心。然后,通过K-means聚类算法将数据集聚类成k个簇,计算数据点在每个维度上的目标函数值并进行升序排列。接着,确定数据点的每个维度的离散程度并选择适当的拟合函数和拟合点,通过最小二乘法对升序排列的每个簇的每1维目标函数值进行函数拟合并求导,以获取变化率。最后,结合信息熵,将每个数据点的每个维度目标函数值乘以相应的变化率进行加权,得到最终的异常得分,并将异常值得分较高的top-n个数据点视为离群点。通过人工数据集和UCI数据集,对KLOD、LOF和KNN方法在准确度上进行仿真实验对比。结果表明KLOD方法相较于KNN和LOF方法具有更高的准确度。本文提出的KLOD方法能够有效改善K-means聚类算法的聚类效果,并且在局部离群点检测方面具有较好的精度和性能。

Low Complexity Minimum Mean Square Error Channel Estimation for Adaptive Coding and Modulation Systems

Performance of the Adaptive Coding and Modulation(ACM) strongly depends on the retrieved Channel State Information(CSI),which can be obtained using the channel estimation techniques relying on pilot symbol transmission.Earlier analysis of methods of pilot-aided channel estimation for ACM systems were relatively little.In this paper,we investigate the performance of CSI prediction using the Minimum Mean Square Error(MMSE)channel estimator for an ACM system.To solve the two problems of MMSE:high computational operations and oversimplified assumption,we then propose the Low-Complexity schemes(LC-MMSE and Recursion LC-MMSE(R-LC-MMSE)).Computational complexity and Mean Square Error(MSE) are presented to evaluate the efficiency of the proposed algorithm.Both analysis and numerical results show that LC-MMSE performs close to the wellknown MMSE estimator with much lower complexity and R-LC-MMSE improves the application of MMSE estimation to specific circumstances.

Wavelet Density Estimation of Censoring Data and Evaluate of Mean Integral Square Error with Convergence Ratio and Empirical Distribution of Given Estimator

Wavelet has rapid development in the current mathematics new areas. It also has a double meaning of theory and application. In signal and image compression, signal analysis, engineering technology has a wide range of applications. In this paper, we use wavelet method, for estimating the density function for censoring data. We evaluate the mean integrated squared error, convergence ratio of given estimator. Also, we obtain empirical distribution of given estimator and verify the conclusion by two simulation examples.

Robust state of charge estimation of lithium-ion battery via mixture kernel mean p-power error loss LSTM with heap-based-optimizer

The state of charge(SOC)estimation of lithium-ion battery is an important function in the battery management system(BMS)of electric vehicles.The long short term memory(LSTM)model can be employed for SOC estimation,which is capable of estimating the future changing states of a nonlinear system.Since the BMS usually works under complicated operating conditions,i.e the real measurement data used for model training may be corrupted by non-Gaussian noise,and thus the performance of the original LSTM with the mean square error(MSE)loss may deteriorate.Therefore,a novel LSTM with mixture kernel mean p-power error(MKMPE)loss,called MKMPE-LSTM,is developed by using the MKMPE loss to replace the MSE as the learning criterion in LSTM framework,which can achieve robust SOC estimation under the measurement data contaminated with non-Gaussian noises(or outliers)because of the MKMPE containing the p-order moments of the error distribution.In addition,a meta-heuristic algorithm,called heap-based-optimizer(HBO),is employed to optimize the hyper-parameters(mainly including learning rate,number of hidden layer neuron and value of p in MKMPE)of the proposed MKMPE-LSTM model to further improve its flexibility and generalization performance,and a novel hybrid model(HBO-MKMPE-LSTM)is established for SOC estimation under non-Gaussian noise cases.Finally,several tests are performed under various cases through a benchmark to evaluate the performance of the proposed HBO-MKMPE-LSTM model,and the results demonstrate that the proposed hybrid method can provide a good robustness and accuracy under different non-Gaussian measurement noises,and the SOC estimation results in terms of mean square error(MSE),root MSE(RMSE),mean absolute relative error(MARE),and determination coefficient R2are less than 0.05%,3%,3%,and above 99.8%at 25℃,respectively.

Efficient Mean Estimation in Log-normal Linear Models with First-order Correlated Errors

In this paper, we propose a log-normal linear model whose errors are first-order correlated, and suggest a two-stage method for the efficient estimation of the conditional mean of the response variable at the original scale. We obtain two estimators which minimize the asymptotic mean squared error （MM） and the asymptotic bias （MB）, respectively. Both the estimators are very easy to implement, and simulation studies show that they are perform better.

THE LAW OF THE ITERATED LOGARITHM OF RANDOM WEIGHTING APPROXIMATION FOR MEAN ERROR──NON．I．I．D．SITUATION

For the dislribulion if mean error under independent but not identicallydislribuled conditions. its approximating dislribution whose precision reachO is obtained.

The Relationship between Deterministic and Ensemble Mean Forecast Errors Revealed by Global and Local Attractor Radii

It has been demonstrated that ensemble mean forecasts, in the context of the sample mean, have higher forecasting skill than deterministic(or single) forecasts. However, few studies have focused on quantifying the relationship between their forecast errors, especially in individual prediction cases. Clarification of the characteristics of deterministic and ensemble mean forecasts from the perspective of attractors of dynamical systems has also rarely been involved. In this paper, two attractor statistics—namely, the global and local attractor radii(GAR and LAR, respectively)—are applied to reveal the relationship between deterministic and ensemble mean forecast errors. The practical forecast experiments are implemented in a perfect model scenario with the Lorenz96 model as the numerical results for verification. The sample mean errors of deterministic and ensemble mean forecasts can be expressed by GAR and LAR, respectively, and their ratio is found to approach2^(1/2) with lead time. Meanwhile, the LAR can provide the expected ratio of the ensemble mean and deterministic forecast errors in individual cases.

基于优化模糊C-means算法的不平衡大数据分类研究

针对不平衡大数据的分类问题,提出一种优化模糊C-means算法的不平衡大数据分类算法。先计算C-means模糊交叉算子,定义优化函数,并求解大数据不平衡增益。利用Spark分类平台,确定大数据样本压缩模糊近邻值的取值范围,再通过放大近邻值的处理方式,定义不平衡阈向量,从而完善整个分类流程,完成基于优化模糊C-means算法的不平衡大数据分类方法的设计。实验结果表明,上述分类方法的应用,可将正例信息、负例信息的取样长度区间完全分离开来,能有效解决因不平衡大数据分类不精准造成的信息样本混淆的问题,符合实际应用需求。

LTE系统中的Mean-OTDOA定位算法

由于LTE蜂窝网中远近效应的影响,终端测量到的邻近基站信号的定位参数会存在较大的偏差,导致OTDOA定位方法(到达时间差定位法)估计的终端位置存在较大误差。基于此,提出一种改进的Mean-OTDOA定位算法。首先估计终端与各基站的时延,然后对终端与多基站的距离测量值进行平均,作为OTDOA定位方法中的参考距离,最后利用泰勒级数展开法对终端位置进行估计。仿真结果表明,该算法可提高终端的定位精度,在基站数目为5、测量误差标准差为50 m时,本算法的均方根误差比OTDOA算法降低了5.2039 m,且随着基站数目的增加,定位精度的改善程度优于OTDOA算法。

基于Fisher线性判别率的加权K-means聚类算法

为提高K-means聚类效果,采用Fisher线性判别率的方法确定特征在聚类中的贡献度并依此对特征进行加权聚类。在人工和实际数据集上所做的实验表明,本方法在聚类效果上优于其他同类加权K-means聚类算法。

一种基于熵和均方差法综合赋权的K-means算法

在传统的K-means聚类算法基础上,本文提出一种基于熵和均方差法综合赋权的Syn-K-means算法。引入综合权重提高聚类结果的类内相似度,从而提高聚类精度。算法中特征权重的计算基于概率论中数字特征的基本描述方法——均方差和信息论中信息特征的基本度量方法——熵;综合赋权系数的选择采用主观设定法求解。实验结果表明,Syn-K-means算法在聚类精度方面优于标准的K-means算法。

改进k-means聚类算法多模型建模的一种新的评价函数

The modeling and control of pH neutralization processes is a difficult problem in the field of process control.A multi-modeling method using an improved k-means clustering based on a new validity function is proposed in this paper.There are some common problems, including the number of clusters assumed as a priori knowledge and initial cluster centers selected randomly for classical k-means clustering.The proposed algorithm is used to compute initial cluster centers and a new validity function is added to determine the appropriate number of clusters, then partial least squares (PLS) is used to construct the regression equation for each local cluster.Simulation results showed that multiple models using the proposed algorithm gave good performance, and the feasibility and validity of the proposed algorithm was verified.

基于改进的Mean Shift鲁棒跟踪算法

Mean Shift跟踪算法在目标尺度变化大和被遮挡时存在较大的缺陷。针对这一问题,提出了一种基于多级正方形匹配的自适应带宽选择和分块抗遮挡的目标跟踪算法。该算法采用目标中心点的离散程度和增量试探法计算出可能的变化尺度,然后采用多级正方形匹配法预测目标的运动趋势,将巴氏系数最大者的尺度作为Mean Shift核函数新的带宽。同时,对前景目标进行分块,根据子块的遮挡程度自适应改变子块权重并按一定准则融合有效子块的跟踪结果。实验结果表明,该算法具有很好的鲁棒性。

一种基于均方差属性加权的K-means算法

在传统的K-means聚类算法基础上提出了一种基于均方差属性加权的MWS-K-means算法。引入特征权重以提高聚类结果的类内相似度(intra-similarities),从而提高聚类精度。考虑到K-means算法采用误差平方和作为聚类准则函数,而误差平方和与概率论中数字特征的基本描述方法———均方差具有较高相似性,算法中特征权重的计算采用均方差法。根据属性的离散程度对欧氏距离进行加权处理,从而用相对距离代替绝对距离来计算类间相似度。实验结果表明:MWS-K-means算法在聚类精度方面优于标准的K-means算法。

基于改进型K-means聚类的温度插值算法

针对K-means聚类算法对初始聚类中心敏感和易陷入局部最优解的缺点及初始聚类中心对聚类结果的影响,提出一种基于改进型K-means聚类和正交最小二乘法的RBFNN算法。利用改进型K-means聚类算法对输入样本数据进行聚类处理,自适应地确定RBFNN隐含层的初始参数,利用正交最小二乘法求隐含层权值,建立RBFNN温度空间插值模型,用已有温度数据加以验证。实验结果表明,该算法能够解决K-means聚类算法对初始聚类中心敏感和易陷入局部最优解的问题,具有较高的插值精度。

基于K-means聚类的数字半色调算法

数字半色调是在二值设备或多色二值设备上实现图像再现的一门技术,提出将K-means聚类法应用在数字半色调技术中。算法中应用人类视觉系统模型(HVS)和印刷模型最大限度减少原始灰度连续调图像和半色调图像之间的视觉误差;利用K-means聚类法将灰度图像划分成聚类分区,在每个聚类分区应用最小平方法(least-squares)最小化二值半色调图像和原始灰度级图像之间的平方误差,所构造的半色调算法与基于模型的最小平方法(LSMB)算法相比,随着聚类分区的增加,图像平滑且边缘清晰度增加,尤其是在图像细节部位。与LSMB算法比较,该算法的均方误差值有所降低,而权重信噪比和峰值信噪比提高了0.2～2 dB,模拟实验结果验证了算法的有效性。

基于有效特征筛选的Mean shift运动目标跟踪算法

针对现有Mean shift跟踪算法在目标被遮挡、跟踪场景变化时,跟踪误差变大甚至丢失目标的问题,提出了一种基于有效特征筛选的Mean shift运动目标跟踪算法。首先通过对目标特征的优化筛选,改善了现有Mean shift算法因目标特征多而造成计算时间较长,在目标发生较大变化时跟踪精度降低的情况。更能有效地表征目标特征,减少跟踪误差,增强特征集对目标的描述能力。同时给出目标模板更新的方法,在目标发生明显变化时,能自适应地更新特征集,进一步提高跟踪精度。仿真结果表明:文中方法具有更好的跟踪精度,计算时间较小,对遮挡、场景变化有更好的鲁棒性。

基于Mean-shift跟踪算法的汽车运行物体目标监测分析

为了提高汽车对前方运动物体的识别能力,开展基于Mean-shift跟踪算法的汽车运行物体目标监测分析。进行双目标定时,将相机内参数代入MATLAB软件中来完成双目标定的过程,获得相机位置数据。完成标定之后,再校正图像。现场测试得到,计算出目标点和汽车之间的位置测试距离和设定距离基本一致,最大误差不超过5%。当距离增大后,目标点更易在左右图像内完成特征点匹配,实现误差降低的目标,并在距离等于5m处达到最小的误差。采用所提算法可以精确跟踪运动目标,并且显示即使物体受到遮挡的情况下也可以完成跟踪的目标,表明算法是可靠的。

基于TWH与FB-error约束的均值漂移无人机跟踪算法

针对无人机跟踪过程中目标遮挡和目标背景变化等因素导致的跟踪失败现象,提出一种M TF(M ean-shift by TWH and FB-error)跟踪算法。首先,在M ean-shift跟踪框架下引入目标加权直方图(TWH:Target-Weighted Histogram)描述目标,即在跟踪过程中,用目标的局部背景来削弱所有区域的内部背景特征,使目标特征突出;其次,添加FB-error约束,在目标被部分遮挡时,通过使用FB-error相关加权函数把目标当前位置的预测结果与Mean-shift矢量计算出的位置结果联合起来估计目标在t时刻的最终位置。实验表明,此跟踪算法在跟踪精度上有较大突破。