主成分分析

万有引力单目标优化算法，用于单目标线性规划问题(Universal gravitation single objective optimization algorithm is used for single target linear programming problem.)

提供函数能进行1维2维的FFT，输入输出的序列长度应为2的整数次方倍(Provides functions can be one-dimensional two-dimensional FFT, sequence length input and output should be an integer power of 2 times)

用LU和LDU分解求解方程组，给出了详细说明(The equations are solved by the decomposition of LU and LDU, and the detailed description is given.)

cressman格点插到格点程序，从一个场插值出另一个分辨率不同的场(interpolates fields to another grid as specified )

C-means in 2 types. C=2

本程序用于实现快速傅里叶变换的快速算法 ％ 为代码简便所有*2操作均直接实现，实际应采用二进制左移实现 (This procedure used to realize fast Fourier transform fast algorithm of all the code is simple to operate* 2 are referred directly to realize the actual binary left should be used to achieve)

用追赶法来求解三对角线性方程组，解决数值分析问题，与大家共同学习。并且提供黄金分割法、高斯赛德尔迭代法求解数值分析问题(The catch-up method is used to solve tridiagonal linear equations, to solve numerical analysis problems, and to learn from you. The Golden Section Method and Gauss Seidel Iteration Method are also provided to solve numerical analysis problems.)

SIRT方法解线性方程组，通过迭代求解，取得了更为精确的结果，节省了计算时间(sirt mehtod to solve equation)

用matlab仿真，通过不同的方法计算矩阵的SVD分解，并用SVD的方法计算最小二乘问题和进行图像压缩(Matlab simulation, different ways to calculate the matrix SVD decomposition and the SVD method to calculate the least squares problem and the image compression)

一种流形学习算法，用于非线性降维，文章发表在2000年science杂志上，是一种非常经典的算法。(A manifold learning algorithm for nonlinear dimensionality reduction, articles published in science journal in 2000, is a very classic algorithms.)