codebar

说明：  双门限算法，用于语音端点检测。可以通过调整门限值，并加入门限自适应算法，实现语音端点检测。(double door ,speech endpoint detection two games adjust itself)

详细介绍了机器学习中非监督学习的内容、分类和方法(This document introduces the unsupervised learing of Machine learning with details.)

去斜率（strech）参考信号与回波信号混频后得到中频输出信号S(t)=Aexp(j*2*pi*f0*t+phase) 其中f0与回波延时的关系为： f0=k*delay_t K为线性调频斜率，delay_t为回波延时 将s(t)采样，FFT变换即可得到目标距离与频谱的关系。 Strech 信号测距 输入混频后的中频信号（图一），经采样及FFT变换得到频谱（图二） 再根据频谱与距离的关系式R=Tcf/2B显示出距离(图三) (Go slope (strech) obtained after the reference signal and the echo signal mixer IF output signal S (t) = Aexp (j* 2* pi* f0* t+phase) f0 and echo delay relationship: f0 = k* delay_t K chirp slope, delay_t the echo delay sampling s (t), the FFT transform to the target distance spectrum of relations. Strech signal ranging input mixer IF signal (Figure 1), sampling and FFT transform spectrum (Figure 2) according the the spectrum distance relationship R = Tcf/2B shows the distance (Figure 3))

介绍了非高斯Levy噪声下的分数阶扩展卡尔曼滤波算法，对非线性离散系统进行滤波。(The fractional order Extended Kalman Filter algorithm for nonlinear discrete system under non-Gaussian Levy noises)

利用这个小数据量方法计算混沌时间序列的最大李雅普诺夫指数(With small amount of data calculated largest Lyapunov exponent of chaotic time series)

滤波求和方式实现宽带波束形成，用MATLAB实现动态聚类或迭代自组织数据分析，匹配追踪和正交匹配追踪。( Filtering summation way broadband beamforming, Using MATLAB dynamic clustering or iterative self-organizing data analysis, Matching Pursuit and orthogonal matching pursuit.)

用二阶自回归（AR2）模型估计信号的瞬时频率(Using second-order autoregressive (AR2) model estimates the instantaneous frequency signal)

这个程序可以演示关于哲学家就餐的问题，对于理解线程的相互等待有很好的作用(This procedure can be demonstrated on the Dining Philosophers problem, each thread to wait for understanding the role of good)

i have written the code for linear classifier for n classes.

一个ＭＡＴＬＡＢ实用例子 一个ＭＡＴＬＡＢ实用例子(a practical example of a practical example of MATLAB)