NMF-matlab

讲UPF的一个PPT,只对一篇经典文章的总结,我觉得很好,比看文章更能抓住重点.(UPF speaking of a PPT, is only a summary of the classic article, I feel good, better than to see the article focused.)

说明：  数字信号处理在电话语音拨号系统中的应用（自动拨号）(Digital signal processing in a telephone voice dialing system (automatic dialing))

实现卷积编码、解码，采用篱笆图的方式实现(implement of juanji)

这是国外做得一个课题，后面附有源代码，matlab的代码(This is done abroad, a task attached to the source code, matlab code)

压缩感知（compressed sensing，cs）在MRI上的应用，该程序生成傅里叶变换域的射线采样模式用作随机测量 (Compressed sensing (compressed sensing, cs) on MRI, the program generates the rays of the Fourier transform domain sampling mode as random measurement)

利用matlab设计高通、低通、带通和带阻滤波器，生成效果图(High-pass, low-pass, band-pass and band-stop filter using matlab design generated effect diagram)

自己写的频率谱细化程序，应用了两种方法实现，经过测试效果不错，拿来和大家分享。(there are two programs that make the Zoom_fft come true.)

一个EKF的简单应用,实现了用EKF算法对三维非线性系统进行仿真。(A simple application of EKF,I used the EKF algorithm to simulate a three-dimensional nonlinear system.)

lingjian.m-----蒙特卡罗方法 lingjian.m使用零件初始值，用蒙特卡罗方法算出总费用。其中使用了自己编制的正态分布随机数发生器产生正态分布随机数。lingjian.m是对蒙特卡罗方法的一次练习。 accyouhua为标定值的函数，而lingjian不是一个函数，在其中已给出了一组标定值的值。 退火确定标定值/unitanneal()----模拟退火 连续型多个变量组合优化问题 这是对模拟退火方法的一次练习，结果证明模拟退火确实是一个行之有效的方法。 当参数选择较好时（一般也伴随着运行时间的加长），模拟退火的结果较好，然而用MATLAB的FMIMCON()一般可达到更高的精度。(lingjian.m---- Monte Carlo method. 氀椀渀最樀椀愀渀.m Part initial value, using the Monte Carlo method to calculate the total cost. The preparation of their own normal distribution random number generator to generate normally distributed random numbers. lingjian.m is the first practice of the Monte Carlo method. accyouhua calibration value the function lingjian not a function, which gives the value of a set of calibration values. Annealing to determine the calibrated value/unitanneal ()---- Simulated Annealing 吀栀攀 continuous multiple variables combinatorial optimization problems 吀栀椀猀 is an exercise of the simulated annealing method, the results show that the simulated annealing is an effective method. 圀栀攀渀 the parameter selection is better (generally accompanied by a longer running time), simulated annealing results, however using MATLAB FMIMCON () generally achieve higher accuracy.)

KNN, K最邻近算法 交叉验证，选出KNN算法中较优的K值(KNN：K nearest neighbours cross validation, for the choice of the best K in KNN algorithm)