stft

利用遗传算法优化RBF神经网络的结构，包括优化权值、高斯基的中心和宽度(The use of genetic algorithm to optimize the structure of RBF neural network, including the right to optimize the value of the Gaussian center and width)

说明：  一个关于傅里叶变换的学习课件，本人自己制作，课题组内部学习使用。(A study on the Fourier transform of the courseware, I produce their own, learning to use the internal research group.)

基于阈值的脑磁共振图像分割，其中的所有代码。(Brain magnetic resonance image segmentation based on the threshold, all code.)

Simple binary tree class

说明：  EKF仅仅利用了非线性函数Taylor展开式的一阶偏导部分（忽略高阶项），常常导致在状态的后验分布的估计上产生较大的误差，影响滤波算法的性能，从而影响整个跟踪系统的性能。最近，在自适应滤波领域又出现了新的算法——无味变换Kalman滤波器（Unscented Kalman Filter-UKF）。UKF的思想不同于EKF滤波，它通过设计少量的σ点，由σ点经由非线性函数的传播，计算出随机向量一、二阶统计特性的传播。因此它比EKF滤波能更好地迫近状态方程的非线性特性，从而比EKF滤波具有更高的估计精度。 (EKF only uses non-linear function of the first-order Taylor expansion of some partial derivatives (ignoring higher order terms), often leading to the posterior distribution of the state estimates to generate large errors affect the performance of filtering algorithms, which affect the whole tracking system performance. Recently, the field of adaptive filtering algorithms and the emergence of new- and tasteless transform Kalman filter (Unscented Kalman Filter-UKF). EKF UKF filter is different from the idea that it points through the design of a small amount of σ by σ point spread through the nonlinear function to calculate the random vector first and second order statistical properties of the transmission. Therefore it is better than the EKF filter nonlinear characteristics equation of state approach, which is more than the EKF filter estimation accuracy.)

说明：  matlab code to solve la16

用于2维和3维的TSP问题，固定起点和终点，可给定旅行商数量(For the 2 dimensional and 3 dimensional TSP problem, fixed starting point and end point, the number of traveling salesmen can be given.)

Elman neural network prediction Mackey-Glass serie

EEGLAB 中文帮助文档，很有用的， 有需要的同学可以下载下来看看(EEGLAB Chinese help documentation, very useful, there is a need for students to be able to download it and see)

wave simlation matlb example