Mean_shift

用三次样条曲线进行精密星历插值和卫星钟差插值(Cubic spline curve the precise ephemeris interpolation and satellite clock error interpolation)

Adaptive noise filter

一个主成分分析方法的matlab实现，比较经典齐全的源码(A matlab Implementation of principal component analysis)

MATLAB小程序，COPULA函数对计算多元统计量的联合概率密度！(MATLAB small programme, which is used to calculate the Copula Function.)

说明：  单变量丢翻图方程程序的M文件,直接运行即可(diophtine)

采用logistic构造的混沌模型，可以产生混沌序列，简单修改参数后可以产生任意区间的混沌序列程序(using logistic structure of chaotic model can produce chaotic sequence, After the simple parameter changes can have arbitrary interval of chaotic sequence procedures)

同时采用身高和体重数据作为特征，用Fisher线性判别方法求分类器，将该分类器应用到训练和测试样本，考察训练和测试错误情况。将训练样本和求得的决策边界画到图上，同时把以往用Bayes方法求得的分类器也画到图上，比较结果的异同。(At the same time the height and weight data as features, Fisher with linear identifying method for classifier, will this classifier applied to training and test sample, to investigate the training and test error conditions. Will the training sample and values of the decision on the graph shows the boundary, and at the same time before the Bayes method with optimal classifier also paint chart, the similarities and differences of the comparison results. )

【谷速软件】卷积的GUI 可以作为参考使用学习([Valley] GUI convolution speed software can use as a reference to learn)

基于“脉动”结构 求 任意偶数阶实对称阵的特征值，用于MIMO通信反馈信道质量，已经在实际产品实现。( 注：脉动结构只需O(n*log(n))量级运算，高效快速)("Systolic Arrays" structure based Eigen Value Decomposition)

The shuffled complex evolution with principal components analysis–University of California at Irvine (SP-UCI) method is a global optimization algorithm designed for high-dimensional and complex problems. It is based on the Shuffled Complex Evolution (SCE-UA) Method (by Dr. Qingyun Duan et al.), but solves a serious problem in searching over high-dimensional spaces," population degeneration". The population degeneration problem refers to the phenomenon that, when searching over the highdimensional parameter spaces, the population of the searching particles is very likely to collapse into a subspace of the parameter space, therefore losing the capability of exploring the entire parameter space. In addition, the SP-UCI method also combines the strength of shuffled complex, the Nelder-Mead simplex, and mutinormal resampling to achieve efficient and effective high-dimensional optimization.