subspace

SVM算法的示例程序,帮助读者理解SVM算法的使用(the use of SVU，to help reader undstander how to user this methods)

基于暗原色去雾的图像代码，用了软抠图和另两种改进算法，附加txt说明文档(Based on dark colors to fog image code, with a soft matting, and another two improved algorithms, additional documentation txt)

分别利用 Lagrange插值法、分段线性插值法与三次样条法进行插值逼近(Respectively, by Lagrange interpolation, piecewise linear interpolation and cubic spline interpolation approximation method)

说明：  小波变换在多尺度下检测分解重构图像，以及融合边缘算法。(Under the multi-scale wavelet decomposition of the reconstructed image detection, and edge fusion algorithm.)

使用Matlab写的EZW算法，里面有实例 (ezw algorithm, there are examples)

一个参照葛德彪书编写的3D FDTD源文件，用matlab语言编写，非常详细的注释，很易看懂,2mur bianjie(GE De-biao a reference book, prepared by 3D FDTD source files, using matlab language, very detailed notes, it is easy to understand, 2mur bianjie)

说明：  该程序包含有matlab相关的工具箱，并提供了 相关的实现算例，可以为初学者使用(contains the procedures related to the Matlab toolbox, and provide relevant examples to achieve, for beginners)

error diffusion2 in matlab

幂法是一种计算矩阵主特征值（矩阵按模最大的特征值）及对应特征向量的迭代方法，特别适用于大型稀疏矩阵。 但是，一般幂法迭代向量v的各个不等于零的分量将随k 趋向于无穷大而使计算机溢出。因此，我们必须对某通幕法进行规范。即规范化幂法 (Power Method is a calculation of the main eigenvalue matrix (matrix according to the largest eigenvalue modulus) and the corresponding eigenvector of the iterative method, especially for large sparse matrix. However, the general power-law iteration vector v is not equal to zero all the components will be as k tends to infinity overflow of the computer. Therefore, we must pass a law to regulate screen. That is, standardized Power Method)

modulated_coded_symbols