(3)PDF

寻找最短路径的程序,希望对大家有用,谢谢!主要是为了完成注册(Finds all pair shortest path.)

用esprit算法对波达方向（DOA）信号进行DOA估计的matlab程序。(DOA (DOA) signal DOA estimation algorithm matlab procedures used esprit.)

实现lle和pca的matlab编程，主要目的是比较lle和pca这两个高维数据降维方法的性能，各有所长。(Use matlab to realise lle and pca ，the purpose is to compare lle and pca，which reduce high dimension data to low data dimension，and each has advantage！)

老外写的小波变换的工具箱，基于MATLAB环境；压缩包内有详细的说明和相关Demo，参考价值较高。(Written by foreigners Wavelet Toolbox, MATLAB-based environment compressed packet has detailed instructions and Demo, a high reference value.)

对文献《Simulation Models With Correct Statistical Properties for Rayleigh Fading Channels》提出的Rayleigh信道模型的仿真，仿真了信道的自相关，互相关，以及四阶统计量。仿真结果与文献一致。目前PDF的仿真的效果不好，还未奉上。关于LCR,AFD也没有做。但是总体说来，目前的仿真结果与文献是完全符合的。对应的文献也放到压缩包了。(This matlab code for paper, <Simulation Models With Correct Statistical Properties for Rayleigh Fading Channels>.This papar is very popular in channel estimation simulation.The paper is also upload with the code.)

自己开发的程序，图像的分形维数计算。看到比较实用才发送上来(Their own development process, the image of the fractal dimension calculation. Was sent up to see more practical)

模拟谐波，然后进行分解，通过傅里叶检测谐波所含谐波的种类(Harmonic analysis and harmonic detection )

copula例子 与GARCH相结合，分析几个指标的相关性。(The copula example is combined with GARCH to analyze the correlation of several indicators.)

用matlab生成下列500*500的RGB红绿蓝彩色图像，框内的颜色如图所示（a）假定将该图像转换到HSI空间，用25*25的平均模板模糊H分量图像，再转换回RGB空间，会看到什么结果?(Generated with matlab the following 500* 500 RGB red, green and blue color image, the color box as shown (a) assumes that the image is converted to HSI space, with the average template 25* 25 Fuzzy H component image, and then converted back RGB space, see what results?)

实现ID3算法，在结果中以树表示出来。决策树是对数据进行分类，以此达到预测的目的。该决策树方法先根据训练集数据形成决策树，如果该树不能对所有对象给出正确的分类，那么选择一些例外加入到训练集数据中，重复该过程一直到形成正确的决策集。(ID3 algorithm to achieve, in the results that come out to the tree. Decision tree is to classify the data, thus achieving the purpose of prediction. The decision tree training set of data according to the formation of the first decision tree, if the tree can not give the correct classification of all objects, then select a number of exceptions to the training set data, repeat the process until the correct decision set.)