mmse-ml-vblast

本程序是卡尔曼滤波的源程序包。内含卡尔曼滤波的常用matlab代码(This procedure is Kalman filtering of the source package. Contains the usual Kalman filter matlab code)

《MATLAB实用教程---控制系统仿真及应用》的配套光盘，不但便于读者使用，而且给任课老师提供了生成数量不限的 练习题的功能，便于因人施教。(" MATLAB tutorial--- practical control system simulation and application of" matching discs, readers not only easy to use, but also to the classroom teacher provides a unlimited number of exercises build functional, easy to teach from person to person.)

关于基音周期的提取，倒谱法的改进，利用预测余量(On the pitch period extraction, improved cepstrum method, the use of margin forecast)

Genetic algorithm implementation

以MATLAB 为基础开发的EEGLAB使用手册的中文版。很实用。自留版(MATLAB as the basis for the development of EEGLAB user manual Chinese version. Very practical. Retention Edition)

采用MATLAB 对图像进行加噪，滤波，置乱等方面的处理代码。(photo noise in matlab)

卡尔曼滤波器预测，以及一周卡尔曼滤波器预测(Kalman filter forecast, and a week Kalman filter forecast)

局部均值分解的MATLAB程序，但是使用有一个问题没有解决，就是分解的时候怎么去除骑行波的问题。(Local mean decomposition MATLAB program, but there is a problem using unresolved, is how to remove the problem of decomposition when riding waves.)

在无约束最优化方法中，通过选择搜索方向 k d 而得到的方法主要有四种：最 速下降法，Newton法，共轭方向法，Quasi-Newton (拟牛顿法)。 (Study on the algorithm of several nonlinear conjugate gradient method and the analysis of global convergence)

先归一化输入输出参数，用KMens优化RBF神经网络中心值，然后计算隐节点数据中心间的距离（矩阵），得到各隐节点的扩展常数（宽度），接着用最小二乘法得到各隐节点的输出权值，对数据进行反归一化，并绘图得到神经网络输出与测试集图，同时进行性能评价(First, the input and output parameters are normalized, the central value of the RBF neural network is optimized by KMens, then the distance (matrix) between the data centers of the hidden nodes is calculated, and the extended constant (width) of the hidden nodes is obtained. Then the output weights of the hidden nodes are obtained by the least square method, and the data are back normalized, and the neural network is drawn. Output and test set diagrams, and performance evaluation)