Genetic algorithm to initialize the network weights and threshold parameters imp...

用matlab语言编写模拟退火算法，实现0－1被包问题-using Matlab language simulated annealing algorithm to achieve 0-1 by Kit

多目标进化算法源代码-more evolutionary algorithm source code

这是用实验函数--Ackley函数来检验遗传算法的优越性，其中有点小毛病，就是种群数在20个左右效果最好，一旦增加种群，结果就不大对劲了。忙了好多天，也没调出来，还请各位大侠指点喽-This is the experimental function-- Ackley function to test the advantages of genetic algorithm, with a little fault, is the number of stocks in the 20 best around, once increased stocks, the result is not very comfortable having the. Busier afterwards, it was not stressed out, please Shanhaiguan kind Lackeys

Multi-objective evolutionary algorithm - an application-Multi-objective evolutionary algorithm- an application

自动发贴机器人，可用于多个论坛的哦。能够突破房SPAM系统 。-Automatically post robot can be used in a number of forums Oh. Able to break through the Housing SPAM system.

用matlab编的，一个典型的BP神经网络的例子，相信对你的学习、工作都能起到一个参考的作用-Using matlab series, a typical example of BP neural network, I believe your study, work can play the role of a reference

以黎曼几何为理论依据，基于S.Amari的修正核函数思想提出了两种新的保角变换，用其对核函数进行数据依赖性改进，进一步提高支持向量机分类器泛化能力。以人工非线性分类问题 为对象进行研究，仿真实验结果表明采用新保角映射可以快速显著地改善分类器泛化性能，而且能大幅度地减少支持向量的数目。-Two novel conformal transformations were proposed based on the Riemannian geometry theory and S.Amari’sidea.And the kernel function was modified by the transformation in a data-dependent way.Our experimental results for theartificial nonlinear data set show that the generalization performance of support vector machines classifier is improvedremarkably and the number of support vectors is decreased greatly.

Lin-Kernighan的R-Opt算法。主要思想是CLK，小规模的TSP可以在很短的时间内都到最优解。-Lin-Kernighan algorithm for the R-Opt. The main idea is CLK, small-scale TSP can be in a very short period of time have gone to the optimal solution.

基于神经网络的字母识别系统 -Letters based on neural network recognition system

1、编制程序显示印章图像（24位真彩色位图）； 2、读出位图中每一像素点的（R，G，B）样本值； 3、以RGB其中某两个（或三个）为坐标，取一定数量的图像点为分析样本，分析其坐标系中的分布； 4、采用本章学习的方法找到分类判别函数，对这些样本进行分类；（要求首先将印章与底纹区分，如有可能将印章、底纹、签字区分） 5、将分类后的结果标记到原始图像上，检查其效果。 -1, the preparation procedures showed that the seal image (24-bit true color bitmap) 2, read out each bitmap pixel (R, G, B) sample value 3 to RGB which a two (or three ) for the coordinates, take a certain number of image points for the analysis of samples, analysis of its coordinate system of distribution 4, the method used in this chapter to learn to find discriminant function classification of these samples for classification (required first to distinguish between the seal and Shading, if possible, will seal, Shading, signatures to distinguish between) 5, will be the result of classification markings to the original image to check its effectiveness.