wnninit

统计模式识别工具箱（Statistical Pattern Recognition Toolbox）包含： 1，Analysis of linear discriminant function 2，Feature extraction: Linear Discriminant Analysis 3，Probability distribution estimation and clustering 4，Support Vector and other Kernel Machines( This section should give the reader a quick overview of the methods implemented in STPRtool. • Analysis of linear discriminant function: Perceptron algorithm and multiclass modification. Kozinec’s algorithm. Fisher Linear Discriminant. A collection of known algorithms solving the Generalized Anderson’s Task. • Feature extraction: Linear Discriminant Analysis. Principal Component Analysis (PCA). Kernel PCA. Greedy Kernel PCA. Generalized Discriminant Analysis. • Probability distribution estimation and clustering: Gaussian Mixture Models. Expectation-Maximization algorithm. Minimax probability estimation. K-means clustering. • Support Vector and other Kernel Machines: Sequential Minimal Optimizer (SMO). Matlab Optimization toolbox based algorithms. Interface to the SVMlight software. Decomposition approaches to train the Multi-class SVM classifiers. Multi-class BSVM formulation trained by Kozinec’s algorithm, Mitchell- Demyanov-Molozenov algorithm )

1、验证窗函数 N 变化时，验证其频谱主瓣副瓣比、主瓣宽度的变化。 a)矩形窗函数的 N 变化时，验证其其频谱主瓣副瓣幅度比基本不会发生变化， 而主瓣宽度将会变窄。这说明，当用矩形窗函数设计滤波器时，增大 N 不能使 得阻带衰减减小，但能够减小过渡带。 b)再选取其他的窗如 hamming/hanning 窗，验证当 N 变化时，其频谱主瓣宽 度变化、主瓣副瓣比值变化情况。 2、用窗函数法设计线性相位 FIR 低通，通带截止频率 wp=0.5PI, 阻带截止频率 ws=0.6PI, 阻带衰减不小于 40dB,通带衰减不大于 3dB. a)选取 Hanning,Hamming 窗查看设计出来的 FIR 的过渡带宽和阻带衰减是否 满足要求，二者有什么不同。 b)使用 hamming 窗，将窗长增大 1 倍，设计 FIR。验证同样的窗函数类型 (hamming)，不同窗长度时，设计出来的 FIR 的过渡带宽和阻带衰减都有什么变 化。 3、不使用 Matlab 系统函数得到上一项指标的低通，并与上一项结果比较。 4 用窗函数法设计线性相位 FIR 高通，通带截止频率 wp=0.8PI, 阻带截止频率 ws=0.7PI, 阻带衰减不小于 30dB,通带衰减不大于 3dB.(An actual filter is implemented using the truncated unit impulse response sequence of the ideal filter. For the ideal low pass filter, the unit impulse response h, N, D () is intercepted by a length of N, and the length is The N sequence H (n), when intercepted, guarantees causality and linear phase requirements for the filter. In order to reduce the Gibbs effect, the H (n) window is added, and the appropriate window function is selected to ensure the stopband attenuation and Transition zone requirement. Note that the sidelobe of the window function affects the stopband attenuation of the filter, and the main lobe width affects the filter Transition bandwidth.)

说明：  均匀圆阵的music算法（包括含相关信号）(Circular array MUSIC algorithm (including correlation signal))

说明：  关于用matlab实现的曲线拟合50000个点的程序，不知道对大家有没有用(Matlab on the use of curve fitting to achieve 50,000 points, the procedures, I do not know all of you there is no use)

EXTRACTING speech features using LPC

neural network classification

C program of binary search

对正弦信号进行FFT变换 画频谱图，进行频谱分析，可以计算出信噪比(Sinusoidal signals of paintings transform FFT spectrum chart, spectrum analysis, signal to noise ratio can be calculated)

A very useful matlab package on super-resolution.(Matlab package on super-reso lution.)

该程序可以通过求解指纹的点分布图和块分布图达到指纹识别的目的。(The program can be obtained by solving the point of the fingerprint maps and block maps to the purpose of fingerprint identification.)