Delaunay

自己编写的2维动态模型，运用卡尔曼滤波仿真。(I have written two-dimensional dynamic model, the use of Kalman filter simulation.)

classification and regression tree algorithm 算法c++实现(classification and regression tree algorithm vc++)

信道建模的大牛Laurent Schumacher的MIMO信道MATLAB仿真程序的使用方法说明文件（PDF格式），完整的图像显示和分析，对MIMO研究工作很有帮助。(The result is an IS-95CDMA forward link software simulation package ,which mimics real-time data communication from a basestation to a cellular unit)

加行高斯白噪声下的低密度奇偶校验码的算法，简单的程序，对初学者有用。(Add the line under the Gaussian white noise low-density parity-check codes of the algorithm, a simple procedure, useful for beginners.)

MATLAB上实现的OFDM+AWGN/Rayleigh+BPSK/QPSK/16QAM/64QAM，并分析EbN0 vs BER.(The MATLAB implementation OFDM+AWGN/Rayleigh+BPSK/QPSK/16QAM/64QAM, and analyze EbN0 vs BER.)

说明：  基于MATLAB完成的神经网络源程序,可以利用该程序完成有关神经网络的应用。(MATLAB based on the completion of the neural network source code, you can use the program to complete the neural network applications.)

的风格特让他上部结构后可见很快将很快均可个黄金的身高的风格(rtdddddsvfhhhjkioiuosadsvbnn)

C++中使用fftw库，根据correlation filters的相关知识编写实现，可以是实现图片的相关滤波的功能，在一副图片中找到对应patch在图片中的位置，从而进行detection和tracking的相关改进(C++ fftw library use, according to the knowledge of the preparation to achieve correlation filters, which can be filtered so as to achieve a picture related to the correlation detection and tracking of improvements)

encoder and decoder programme for pre-coding MIMO system based on THP algorithm (encoder and decoder programme for pre-coding MIMO system based on THP algorithm)

这个Matlab程序实现最速下降算法。最速下降法是一种最基本的算法，它在最优化方法中占有重要地位.最速下降法的优点是工作量小，存储储变量较少,初始点要求不高；缺点是收敛慢，最速下降法适用于寻优过程的前期迭代或作为间插步骤，当接近极值点时，宜选用别种收敛快的算法. (The Matlab program to achieve the steepest descent algorithm. Steepest descent method is a basic algorithm, it occupies an important position in the optimization method. Advantage of the steepest descent method is the small amount of work, less storage reservoir variables, the initial point less demanding drawback is slow convergence, the steepest descent method for the optimization process of the early iterations, or as inserted steps are advised to use other kinds of fast convergence of the algorithm when close to the extreme point.)