《遗传算法理论应用与软件实现》中的初始化MATLAB源代码

通过C4.5 的实现可以进行构建决策树 来进行有效的分类-Through the realization of C4.5 decision tree can be constructed to carry out an effective classification

随机神经网络工具箱-random neural network toolbox

采用“提前停止”方法提高 BP 网络的推广能力。对于和例 2相同的问题，在本例中我们将采用训练函数 traingdx -The use of

专家浅谈人工智能，doc文档-expert on artificial intelligence, doc file

基于Matlab的简单遗传算法和简单bp网络的程序实现-based on Matlab simple genetic algorithms and simple bp Network Program

一个语音信号端点检测的程序，算法可能需要根据具体的应用调整一些参数。请使用者自己调整，根据不同的噪声环境。-A voice signal endpoint detection procedure, the algorithm may need to adjust according to specific application parameters. Please adjust the user, according to different noise environments.

基于canny的灰色预测和模式识别,标定的一种做法是：在一块平板上画若干个点， 同时用这两部相机照相，分别得到这些点在它们像平面上的像点，利用这两组像点的几何关系就可以得到这两部相机的相对位置。然而，无论在物平面或像平面上我们都无法直接得到没有几何尺寸的“点”。实际的做法是在物平面上画若干个圆（称为靶标），它们的圆心就是几何的点了。而它们的像一般会变形，如图1所示，所以必须从靶标上的这些圆的像中把圆心的像精确地找到，标定就可实现。-Based on the gray canny prediction and pattern recognition, calibration approach is to: a plate painted in a number of points, at the same time with these two cameras camera, respectively, as these points in their plane as points, the use of these two groups geometric relationship between the image point can be that the relative positions of two cameras. However, in terms of objects such as plane or plane, we can not directly be no geometry point. Actual practice is complex plane painted in a number of round (known as targets), their geometric center of a circle is the point. And they generally like deformation, as shown in Figure 1, it must target these round, like in the center of a circle of like to find precise, calibration can achieve.

关联规则的频繁项集算法bomo的源代码，含说明文件，C编写，可在linux和windows下编译。-Related and regular frequent source codes which collects algorithm bomo, includes the supporting paper, C is written, can compile under linux and windows

一种通过自组织竞争学习网络实现数据降维和可视化的单层神经网络模型。用此算法可以把输入空间的多维映射到低维的（一维或者二维）的离散网络上，并将保持相同性质的输入数据在映射到低维空间时的拓扑一致性。iris以及letter两个数据集进行分类-A competitive learning through self-organizing network for data dimensionality reduction and visualization of single-layer neural network model. Using this algorithm can be multi-dimensional input space is mapped to the low-dimensional (one-dimensional or two-dimensional) discrete network, and will remain the same as the nature of the input data is mapped to the low-dimensional space of topological consistency. iris as well as the letter the two data sets to classify

HMM用于人脸识别，从OPENCV里提取出来的,非常经典的说-HMM for face recognition, OPENCV extract from the very classical said