svm测试

基于模糊神经技术的智能控制过程,模糊神经控制器（NFC）用于复杂系统的智能控制(Based on Fuzzy Neural Intelligent control process, fuzzy neural controller (NFC) for intelligent control of complex systems)

零相移滤波器m函数，优点是克服了滤波器产生的相位滞后，缺点是本仿真不能保证实时性，其原理为输入一次序列，正反各变换一遍。(Zero phase shift filter m function, the advantages of the filter to overcome the phase lag, the disadvantage is that the simulation can not guarantee real-time, the principle of the input sequence, the pros and cons of each conversion.)

支持向量机的实现程序，用matlab编写(Implementation of support vector machine procedures, written with matlab)

compress2D with matlab

matlab使用文本文件进行数据存取详细过程说明(matlab using a text file for data access detailed process description)

基于matlab实现语音信号的采集和处理，包括频谱分析，显示实时A声级(Based on matlab realize the voice signal acquisition and processing, including spectral analysis, display real-time A sound level)

进行逐步线性回归算法，找到最合适的回归模型，进行数据分析(making the data is easy to look for the answer )

做过模糊PID的朋友很多遇到过这种问题，就是仿真启动不起来，停留在e-7等地方或者仿真很慢，自次上次有个朋友说改变步长能解决这个问题，最后的曲线能达到满足要求的时候，我让他找个例子他就没有后文，以前我做了很多这样的模糊PID，但靠改变步长并没有解决问题，仿真曲线直接发散出去了，我自己这两天也总在思考这个问题，通过不断的调试总结，明白了不少，给遇到这个问题的朋友讲解一下解决办法吧，以一个例子详解：(Done a lot of friends of the fuzzy PID encountered this problem, simulation start up, stay in the e-7 and other local or simulation is very slow, have a friend who said to change the step size to solve this problem since the times the last, the finalcurve to achieve to meet the requirements when I asked him to find examples of his later, I have done a lot of fuzzy PID, but by changing the step does not solve the problem, the simulation curve directly divergence out, Itwo days always thinking about this problem, summarized by the debug and understand a lot of friends to experience this problem to explain the solution, an example Detailed:)

复杂网络的matlab程序，包括bnet，nn net和nw net共8个.m文件组成。(compclex network)

该程序是LWSG_NMF方法（局部加权稀疏图正则非负矩阵分解）的matlab实现，可根据所需参数，进行设置(The program is LWSG_NMF method (locally weighted sparse graph regular NMF) matlab implementation, according to the required parameters, set)