一个牛人的神经网络支持向量机程序,对于搞神经网络的朋友强烈推荐.第二部分...

文中设计了一个3层径向基神经网络(RBFN)用于对企业的5项评价指标进行聚类分析，并与蚁群算法做了比较分析。RBFN由输入层 到隐含层采用传统的K一均值算法，隐含层到输出层通过“模2递减”学习速率的BP学习；蚁群算法根据信息素的分配能够自动调整收索 路径，从而达到数据自动聚类的目的。结果表明，与蚁群算法相比，改进RBFN具有快速收敛、自动识别奇异样本的优点，而蚁群算法 无须教师学习，并能够达到全局最优。-In this paper, we designed a 3-layer RBF neural network (RBFN) for the 5-to-business evaluation indicators cluster analysis and ant colony algorithm has done a comparative analysis. RBFN from input layer to hidden layer using the traditional K-means algorithm, hidden layer to output layer through the Mode 2 decreasing learning rate of BP learning ant colony algorithm based on pheromone can automatically adjust the allocation of land Faso path, thereby to achieve the purpose of automatic data clustering. The results showed that compared with the ant colony algorithm to improve the RBFN has a fast convergence, automatic identification of singular advantage of the sample, while the ant colony algorithm do not need teachers to learn and be able to reach the global optimum.

Pascal Programs Printed in GENETIC ALGORITHMS IN SEARCH, OPTIMIZATION, AND MACHINE LEARNING by David E. Goldberg -Pascal Programs Printed in GENETIC ALGORITHMS IN SEARCH, OPTIMIZATION, AND MACHINE LEARNING by David E. Goldberg

基于神经网络的辨识工具箱 (527KB)-Neural Network Based Identification Toolbox (527KB)

　基于VB的遗传算法软件实现 在程序中,FitnessValue (i) 为适应度值数组、avFit2nessValue (100) 为归一化适应度值数组、Population2 Chrom(i ,j) 为遗传个体的等位基因值、Popsize 为种群中的个体数,CHROMLENGTH为一母体对的等位基因 总数。-VB-based genetic algorithm software implementation in the proceedings, FitnessValue (i) In order to meet an array value, avFit2nessValue (100) for the normalized fitness value of the array, Population2 Chrom (i, j) for the genetic value of individual alleles, Popsize for stocks in the number of individuals, CHROMLENGTH for one parent to the total number of alleles.

这个是用来进行基频检测的，目前做的最好的应该数Paul的基频提取算法了，误差可以小于万分之五。该源代码的算法是我根据Paul的算法修改得到，省略了插值部分，所以误差在百分之五以内，但是在实际工作中已经足够了。-is for the frequency of detection, so the best should be the number of frequency Paul extraction algorithm, can be less than prior error. The source code for the algorithm is based on the algorithm Paul to be amended, omitted the interpolation part, the error within 5%, but the actual work has been sufficient.

聚类分析简介，是个课件，希望对大家有帮助。-Introduction Cluster analysis is a courseware, we want to help.

蚁群算法上传一个文件这么复杂啊？生命在长期进化过程中，积累了很多新奇的功能，人类很早就从中得到启发而改进自己的工具，如史书中记戴“见蓬转而做车辑”，传说鲁班被茅苇划破，而发明锯子……也许早先的发明,只是偶然的模仿和发现,后来人们已有意识地进行这方面的研究,这就是“仿生学”。仿生学顾名思义就是模仿生物的某些功能的学问。有名的例子很多，如模仿海豚皮而构造的“海豚皮游泳衣”、科学家研究鲸鱼的皮肤时，发现其上有沟漕的结构，于是有个科学家就依照鲸鱼皮构造，造成一个薄膜蒙在飞机的表面，据实验可节约能源3%，若全国的飞机都蒙上这样的表面，每年可节约几十亿。又如有科学家研究蜘蛛，发现蜘蛛的腿上没有肌肉，有脚的动物会走，主要是靠肌肉的收缩，现在蜘蛛没有肌肉为什么会走路？经研究蜘蛛不是靠肌肉的收缩进行走路的，而是靠其中的“液压”的结构进行走路，据此人们发明了液压步行机……总之,从自然界得到启迪, 模仿其结构进行发明创造.这就是仿生学. 这是我们向自然界学习的一个方面.另一方面,我们还可以从自然的规律中得到启迪,利用其原理进行设计(包括设计算法),这就是智能计算的思想.-Ant Algorithm upload a document so complicated ah? Life in the long evolutionary process, the accumulated growth of many functions, from very early human to be inspired to improve their own tools, such as history books remember wearing "See Peng turn out of the car series," Mao legend Luban Reed was scratched, and the invention saws ... perhaps earlier invention, only occasionally parody and found that there was awareness of the people to carry out this kind of research, and it is a "bionics."

模拟退火算法工具箱介绍 研究sa很有用的东西-simulated annealing algorithm toolbox the study sa very useful things

遗传算法的四个源程序和源代码，在国内没有过的，看看国外的程序员是怎么编写的-Four genetic algorithm source code and source code, never at home and abroad to see how the programmer is prepared

knn算法的实现，框架十分清楚。是经典的东西-KNN algorithm, the framework is very clear. Is a classic East-West