GA

说明：  摘 要 盲图像恢复的主要困难是信息不足，而为了恢复图像和确定点扩散函数需要适当的先验知识。解决这个问题的法、 法以及正则化方法等。但是这些方法的计算量都太大，针对上述方法的不足，文章提出了一种恢复图方法有图像的新算法，它通过恢复残差的最小化和后验概率的最大化来估计参数和恢复图像。其中，巧妙地利用了最陡梯度法和 共轭梯度法的迭代求解。对由于运动造成的模糊图像，可以明显地改善图像的质量，实验结果证明，在对模糊操作没有严格限制的情况下，仍可得到较好的恢复图像。(： : 2$*/<$ I?E L691/6@ F9MM9G:0=@ R9=? S092F 91/DE 6EC=76/=972 9C 92C:MM9G9E2= 92M761/=972$I?9C G/00C M76 L67LE6 92G76L7 （ ） =972 7M/L69769 U27R0EFDE /S7:= =?E 91/DE /2F =?E L792=8CL6E/F M:2G=972 &[ $RE008U27R CL/GE8/F/L=9OE 6ED:0/69</= 1E=?7F /2F 3+ 1E=?7F M76 91/DE 6EC=76/=972 /6E E =E2FEF =7 /FF6ECC =?9C L67S0E1$] 2ER 1E=?7F 9C L67L7CEF =7 6EC= S0:6 91/DE :C92D+5 /2F 5E/C=V:/6EC 1E=?7F$W= :CE 192919<E/G7C= M:2G=972 /2F 1/ 919<E &[ M:2G=972 =7 6EC= =?E 91/DE /2F 9FE2=9M@ =?E L/6/1E=E6C$I?9C 1E=?7F :CE C=EELEC= FECGE2= /2F G72Q:D/=E D6/F9E2= 1E=?7FC =7 192919<E ， G7C= M:2G=972 /2F 1/ 919<E L7C=E69769 L67S/S909=@ =7 91L67OE =?E S0:6 91/DE X?9G? 9C /0C7 =7 91L67OE =?E 91/DE V: =@ 92 C71E FED6EE R9=? 27 C=69G= 0919=/=972$)

Analysis and Applications of Artificial Neural Networks

Hurst数计算，用于算法交易，非常有用(Hurst index)

LSL自适应算法的MATLAB仿真，包括LSL自适应算法的流程仿真，性能仿真，信号模型的构建(MATLAB simulation of LSL)

Introduction to the Companion MATLAB Codes (Introduction to the Companion MATLAB Codes)

matbal100小程序例子，供初学或急用matlab者分享。(matbal100 small program examples for the beginner or emergency matlab share.)

算术编码的实现 对于初学者来说还是比较好用的 (算术的实现 对于初学者来说还是比较好用的 )

详细介绍了有限差分法，以及其matlab实现(Details of the finite difference method, and its implementation matlab)

A code to solve a 2-D truss structure using FEM theorem (direct method). Each branch of the truss is considered as a 2-force member and the coordinates of the nodes in Cartesian system is needed. This code contains 3 functions: Truss_f, which is the main function to calculate stress, strain and force.M_properties22, which contains the mechanical properties of whole truss in a matrix P_condition22, which contains boundary condition and external forces in a matrix

GARCH 模型计算工具包，计算各种garch模型比如GARCH-M，NGARCH，EGARCH(GARCH toolkil for different GARCH model, include GARCH-M, EGARCH,NGARCH etc)