基于神经网络的数字识别方法

较好的计算阻尼Newton法的C++程序，可以很快得到较精确的结果-Better calculation of the damping Newton method C++ program, you can quickly get more accurate results

解算高阶矩阵，可以求逆，求特征值，对矩阵进行特征值分解-High-order matrix solvers can be inverse, and eigenvalues of matrix eigendecomposition

A matlab program for  multi-target tracking,The problem of multi-target tracking is comprised of two distinct, but tightly coupled challenges: (i) the naturally discrete problem of data association, i.e. assigning image observations to the appropriate target; (ii) the naturally continuous problem of trajectory estimation, i.e. recovering the trajectories of all targets. To go beyond simple greedy solutions for data association, recent approaches often

清华大学2002年出版的《科学与工程数值计算算法VC＋＋》配套源码，包括复数计算、矩阵运算、插值计算、线性非线性方程组求解、数值积分、数值拟和与逼近等算法和类，对于科学和工程数值计算很有参考价值！-Qinghua University, in 2002 published "Science and engineering numerical algorithm VC++" complementary source, includes the plural, matrix computation, interpolation, linear solve nonlinear equations, numerical integration, and the proposed numerical approximation algorithms and other categories, science and engineering numerical computation of great reference value!

此代码用于使7段显示器老化；

此程序通过凸优化中的armigo规则的回溯线搜索方法来寻找函数的最佳值。对于输入，此程序是全局aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa

64位数的计算，是科学计算类的-64 median is calculated kind of scientific computing

对满足weibull分布的风速进行monte carol 模拟，计算结果可以作为风速的概率分布

c语言编写数学常用算法--commonly math algorithm written in C language

一些常用的矩阵操作，包括常用的矩阵加法，乘法，strassen算法，还有位并行操作预留的算法。包括为wingoth算法预留的矩阵乘法，为了分块矩阵预留的算法等等。