主要是实现调制识别，区分几种常用的数字调制信号，包括ASK,FSK,PSK,QAM。含有两个文件夹 其一为特征参数的仿真；其二为正确识别率的仿真。 文件夹key feature simulink中： 运行程序会得到各特征参数之间区分图 从图中可看到特征参数的有效性。 文件夹classification rate simulink中： 运行main.m文件 可以得到正确识别率 -Mainly to achieve modulation recog

用核主成分分析算法进行的人脸识别，总的来说识别率还不错-Face recognition using the nuclear principal component analysis algorithm, on the whole pretty good recognition rate

该代码为基于BP神经网络的四种语音的识别，识别率较高。-The code for four kind of voice recognition based on BP neural network, the recognition rate is higher.

通过反复训练模板能有较高的识别率，关于小波的matlab复合分析，FIR 底通和带通滤波器和IIR 底通和带通滤波器。- Through repeated training pGiScdLlate have higher recognition rate, Matlab wavelet analysis on complex, Bottom-pass and band-pass FIR and IIR filter bottom pass and band-pass filter.